liu.seSök publikationer i DiVA
Ändra sökning
Avgränsa sökresultatet
1234 51 - 100 av 195
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Träffar per sida
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sortering
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
Markera
Maxantalet träffar du kan exportera från sökgränssnittet är 250. Vid större uttag använd dig av utsökningar.
  • 51.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    On the Relation Between Anisotropic Diffusion and Iterated Adaptive Filtering2008Ingår i: Pattern Recognition: 30th DAGM Symposium Munich, Germany, June 10-13, 2008 Proceedings / [ed] Gerhard Rigoll, Springer Berlin/Heidelberg, 2008, 1, , s. 436-445s. 436-445Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we present a novel numerical approximation scheme for anisotropic diffusion which is at the same time a special case of iterated adaptive filtering. By assuming a sufficiently smooth diffusion tensor field, we simplify the divergence term and obtain an evolution equation that is computed from a scalar product of diffusion tensor and the Hessian. We propose further a set of filters to approximate the Hessian on a minimized spatial support. On standard benchmarks, the resulting method performs in average nearly as good as the best known denoising methods from the literature, although it is significantly faster and easier to implement. In a GPU implementation video real-time performance is achieved for moderate noise levels.

  • 52.
    Felsberg, Michael
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Optical flow estimation from monogenic phase.2006Ingår i: International Workshop on Complex Motion,2004, Springer , 2006Konferensbidrag (Refereegranskat)
  • 53.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Probabilistic and biologically inspired feature representations2018Bok (Refereegranskat)
    Abstract [en]

    Under the title "Probabilistic and Biologically Inspired Feature Representations," this text collects a substantial amount of work on the topic of channel representations. Channel representations are a biologically motivated, wavelet-like approach to visual feature descriptors: they are local and compact, they form a computational framework, and the represented information can be reconstructed. The first property is shared with many histogram- and signature-based descriptors, the latter property with the related concept of population codes. In their unique combination of properties, channel representations become a visual Swiss army knife—they can be used for image enhancement, visual object tracking, as 2D and 3D descriptors, and for pose estimation. In the chapters of this text, the framework of channel representations will be introduced and its attributes will be elaborated, as well as further insight into its probabilistic modeling and algorithmic implementation will be given. Channel representations are a useful toolbox to represent visual information for machine learning, as they establish a generic way to compute popular descriptors such as HOG, SIFT, and SHOT. Even in an age of deep learning, they provide a good compromise between hand-designed descriptors and a-priori structureless feature spaces as seen in the layers of deep networks.

  • 54.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Spatio-featural scale-space2009Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    Linear scale-space theory is the fundamental building block for many approaches to image processing like pyramids or scale-selection. However, linear smoothing does not preserve image structures very well and thus non-linear techniques are mostly applied for image enhancement. A different perspective is given in the framework of channel-smoothing, where the feature domain is not considered as a linear space, but it is decomposed into local basis functions. One major drawback is the larger memory requirement for this type of representation, which is avoided if the channel representation is subsampled in the spatial domain. This general type of feature representation is called channel-coded feature map (CCFM) in the literature and a special case using linear channels is the SIFT descriptor. For computing CCFMs the spatial resolution and the feature resolution need to be selected.

    In this paper, we focus on the spatio-featural scale-space from a scale-selection perspective. We propose a coupled scheme for selecting the spatial and the featural scales. The scheme is based on an analysis of lower bounds for the product of uncertainties, which is summarized in a theorem about a spatio-featural uncertainty relation. As a practical application of the derived theory, we reconstruct images from CCFMs with resolutions according to our theory. The results are very similar to the results of non-linear evolution schemes, but our algorithm has the fundamental advantage of being non-iterative. Any level of smoothing can be achieved with about the same computational effort.

  • 55.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Spatio-featural scale-space2009Ingår i: Scale Space and Variational Methods in Computer Vision: Second International Conference, SSVM 2009, Voss, Norway, June 1-5, 2009. Proceedings / [ed] Xue-Cheng Tai, Knut Mørken, Marius Lysaker, Knut-Andreas Lie, Springer Berlin/Heidelberg, 2009, s. 808-819Konferensbidrag (Refereegranskat)
    Abstract [en]

    Linear scale-space theory is the fundamental building block for many approaches to image processing like pyramids or scale-selection. However, linear smoothing does not preserve image structures very well and thus non-linear techniques are mostly applied for image enhancement. A different perspective is given in the framework of channel-smoothing, where the feature domain is not considered as a linear space, but it is decomposed into local basis functions. One major drawback is the larger memory requirement for this type of representation, which is avoided if the channel representation is subsampled in the spatial domain. This general type of feature representation is called channel-coded feature map (CCFM) in the literature and a special case using linear channels is the SIFT descriptor. For computing CCFMs the spatial resolution and the feature resolution need to be selected. In this paper, we focus on the spatio-featural scale-space from a scale-selection perspective. We propose a coupled scheme for selecting the spatial and the featural scales. The scheme is based on an analysis of lower bounds for the product of uncertainties, which is summarized in a theorem about a spatio-featural uncertainty relation. As a practical application of the derived theory, we reconstruct images from CCFMs with resolutions according to our theory. The results are very similar to the results of non-linear evolution schemes, but our algorithm has the fundamental advantage of being non-iterative. Any level of smoothing can be achieved with about the same computational effort.

  • 56.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    The GET Operator2004Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we propose a new operator which combines advantages of monogenic scale-space and Gaussian scale-space, of the monogenic signal and the structure tensor. The gradient energy tensor (GET) defined in this paper is based on Gaussian derivatives up to third order using different scales. These filters are commonly available, separable, and have an optimal uncertainty. The response of this new operator can be used like the monogenic signal to estimate the local amplitude, the local phase, and the local orientation of an image, but it also allows to measure the coherence of image regions as in the case of the structure tensor

  • 57.
    Felsberg, Michael
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Wiener channel smoothing: Robust Wiener filtering of images2005Ingår i: Pattern Recognition, 2005, Vol. 3663, s. 468-475Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper, we combine the well-established technique of Wiener filtering with an efficient method for robust smoothing: channel smoothing. The main parameters to choose in channel smoothing are the number of channels and the averaging filter. Whereas the number of channels has a natural lower bound given by the noise level and should for the sake of speed be as small as possible, the averaging filter is a less obvious choice. Based on the linear behavior of channel smoothing for inlier noise, we derive a Wiener filter applicable for averaging the channels of an image. We show in some experiments that our method compares favorable with established methods.

  • 58.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Kalkan, Sinan
    University of Göttingen.
    Krüger, Norbert
    University of South Denmark.
    Continuous dimensionality characterization of image structures2009Ingår i: Image and Vision Computing, ISSN 0262-8856, E-ISSN 1872-8138, Vol. 27, nr 6, s. 628-636Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Intrinsic dimensionality is a concept introduced by statistics and later used in image processing to measure the dimensionality of a data set. In this paper, we introduce a continuous representation of the intrinsic dimension of an image patch in terms of its local spectrum or, equivalently, its gradient field. By making use of a cone structure and barycentric co-ordinates, we can associate three confidences to the three different ideal cases of intrinsic dimensions corresponding to homogeneous image patches, edge-like structures and junctions. The main novelty of our approach is the representation of confidences as prior probabilities which can be used within a probabilistic framework. To show the potential of our continuous representation, we highlight applications in various contexts such as image structure classification, feature detection and localisation, visual scene statistics and optic flow evaluation.

  • 59.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Berg, Amanda
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten. Termisk Systemteknik AB, Linköping, Sweden.
    Häger, Gustav
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Ahlberg, Jörgen
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten. Termisk Systemteknik AB, Linköping, Sweden.
    Kristan, Matej
    University of Ljubljana, Slovenia.
    Matas, Jiri
    Czech Technical University, Czech Republic.
    Leonardis, Ales
    University of Birmingham, United Kingdom.
    Cehovin, Luka
    University of Ljubljana, Slovenia.
    Fernandez, Gustavo
    Austrian Institute of Technology, Austria.
    Vojır, Tomas
    Czech Technical University, Czech Republic.
    Nebehay, Georg
    Austrian Institute of Technology, Austria.
    Pflugfelder, Roman
    Austrian Institute of Technology, Austria.
    Lukezic, Alan
    University of Ljubljana, Slovenia.
    Garcia-Martin8, Alvaro
    Universidad Autonoma de Madrid, Spain.
    Saffari, Amir
    Affectv, United Kingdom.
    Li, Ang
    Xi’an Jiaotong University.
    Solıs Montero, Andres
    University of Ottawa, Canada.
    Zhao, Baojun
    Beijing Institute of Technology, China.
    Schmid, Cordelia
    INRIA Grenoble Rhˆone-Alpes, France.
    Chen, Dapeng
    Xi’an Jiaotong University.
    Du, Dawei
    University at Albany, USA.
    Shahbaz Khan, Fahad
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Porikli, Fatih
    Australian National University, Australia.
    Zhu, Gao
    Australian National University, Australia.
    Zhu, Guibo
    NLPR, Chinese Academy of Sciences, China.
    Lu, Hanqing
    NLPR, Chinese Academy of Sciences, China.
    Kieritz, Hilke
    Fraunhofer IOSB, Germany.
    Li, Hongdong
    Australian National University, Australia.
    Qi, Honggang
    University at Albany, USA.
    Jeong, Jae-chan
    Electronics and Telecommunications Research Institute, Korea.
    Cho, Jae-il
    Electronics and Telecommunications Research Institute, Korea.
    Lee, Jae-Yeong
    Electronics and Telecommunications Research Institute, Korea.
    Zhu, Jianke
    Zhejiang University, China.
    Li, Jiatong
    University of Technology, Australia.
    Feng, Jiayi
    Institute of Automation, Chinese Academy of Sciences, China.
    Wang, Jinqiao
    NLPR, Chinese Academy of Sciences, China.
    Kim, Ji-Wan
    Electronics and Telecommunications Research Institute, Korea.
    Lang, Jochen
    University of Ottawa, Canada.
    Martinez, Jose M.
    Universidad Aut´onoma de Madrid, Spain.
    Xue, Kai
    INRIA Grenoble Rhˆone-Alpes, France.
    Alahari, Karteek
    INRIA Grenoble Rhˆone-Alpes, France.
    Ma, Liang
    Harbin Engineering University, China.
    Ke, Lipeng
    University at Albany, USA.
    Wen, Longyin
    University at Albany, USA.
    Bertinetto, Luca
    Oxford University, United Kingdom.
    Danelljan, Martin
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Arens, Michael
    Fraunhofer IOSB, Germany.
    Tang, Ming
    Institute of Automation, Chinese Academy of Sciences, China.
    Chang, Ming-Ching
    University at Albany, USA.
    Miksik, Ondrej
    Oxford University, United Kingdom.
    Torr, Philip H S
    Oxford University, United Kingdom.
    Martin-Nieto, Rafael
    Universidad Aut´onoma de Madrid, Spain.
    Laganiere, Robert
    University of Ottawa, Canada.
    Hare, Sam
    Obvious Engineering, United Kingdom.
    Lyu, Siwei
    University at Albany, USA.
    Zhu, Song-Chun
    University of California, USA.
    Becker, Stefan
    Fraunhofer IOSB, Germany.
    Hicks, Stephen L
    Oxford University, United Kingdom.
    Golodetz, Stuart
    Oxford University, United Kingdom.
    Choi, Sunglok
    Electronics and Telecommunications Research Institute, Korea.
    Wu, Tianfu
    University of California, USA.
    Hubner, Wolfgang
    Fraunhofer IOSB, Germany.
    Zhao, Xu
    Institute of Automation, Chinese Academy of Sciences, China.
    Hua, Yang
    INRIA Grenoble Rhˆone-Alpes, France.
    Li, Yang
    Zhejiang University, China.
    Lu, Yang
    University of California, USA.
    Li, Yuezun
    University at Albany, USA.
    Yuan, Zejian
    Xi’an Jiaotong University.
    Hong, Zhibin
    University of Technology, Australia.
    The Thermal Infrared Visual Object Tracking VOT-TIR2015 Challenge Results2015Ingår i: Proceedings of the IEEE International Conference on Computer Vision, Institute of Electrical and Electronics Engineers (IEEE), 2015, s. 639-651Konferensbidrag (Refereegranskat)
    Abstract [en]

    The Thermal Infrared Visual Object Tracking challenge 2015, VOTTIR2015, aims at comparing short-term single-object visual trackers that work on thermal infrared (TIR) sequences and do not apply prelearned models of object appearance. VOT-TIR2015 is the first benchmark on short-term tracking in TIR sequences. Results of 24 trackers are presented. For each participating tracker, a short description is provided in the appendix. The VOT-TIR2015 challenge is based on the VOT2013 challenge, but introduces the following novelties: (i) the newly collected LTIR (Linköping TIR) dataset is used, (ii) the VOT2013 attributes are adapted to TIR data, (iii) the evaluation is performed using insights gained during VOT2013 and VOT2014 and is similar to VOT2015.

  • 60.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Duits, R
    Florack, L
    The monogenic scale space on a bounded domain and its applications2003Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we present a method to implement the monogenic scale space on a bounded domain and show some applications. The monogenic scale space is a vector valued scale space based on the Poisson scale space, which establishes a sophisticated alternative to the Gaussian scale space. The features of the monogenic scale space, including local amplitude, local phase, local orientation, local frequency, and phase congruency, are much easier to interpret in terms of image features evolving through scale than in the Gaussian case. Furthermore, applying results from harmonic analysis, relations between the features are obtained which improve the understanding of image analysis. As applications, we present a very simple but still accurate approach to image reconstruction from local amplitude and local phase and a method for extracting the evolution of lines and edges through scale.

  • 61.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Duits, R.
    University of Technology, Eindhoven.
    Florack, L.
    University of Technology, Eindhoven.
    The Monogenic Scale Space on a Rectangular Domain and its Features2005Ingår i: International Journal of Computer Vision, ISSN 0920-5691, E-ISSN 1573-1405, Vol. 64, nr 2--3Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    In this paper we present a novel method to implement the monogenic scale space on a rectangular domain. The monogenic scale space is a vector valued scale space based on the Poisson scale space, which establishes a sophisticated alternative to the Gaussian scale space. Previous implementations of the monogenic scale space are Fourier transform based, and therefore suffer from the implicit periodicity in case of finite domains. The features of the monogenic scale space, including local amplitude, local phase, local orientation, local frequency, and phase congruency, are much easier to interpret in terms of image features evolving through scale than in the Gaussian case. Furthermore, applying results from harmonic analysis, relations between the features are obtained which improve the understanding of image analysis. As applications, we present a very simple but still accurate approach to image reconstruction from local amplitude and local phase and a method for extracting the evolution of lines and edges through scale.

  • 62.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Forssen, P.-E.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Datorseende.
    Scharr, H.
    IEEE Computer Society, Forschungszentrum Jülich GmbH, ICG-III, 52425 Jülich, Germany.
    Channel smoothing: Efficient robust smoothing of low-level signal features2006Ingår i: IEEE Transaction on Pattern Analysis and Machine Intelligence, ISSN 0162-8828, E-ISSN 1939-3539, Vol. 28, nr 2, s. 209-222Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    In this paper, we present a new and efficient method to implement robust smoothing of low-level signal features: B-spline channel smoothing. This method consists of three steps: encoding of the signal features into channels, averaging of the channels, and decoding of the channels. We show that linear smoothing of channels is equivalent to robust smoothing of the signal features if we make use of quadratic B-splines to generate the channels. The linear decoding from B-spline channels allows the derivation of a robust error norm, which is very similar to Tukey's biweight error norm. We compare channel smoothing with three other robust smoothing techniques: nonlinear diffusion, bilateral filtering, and mean-shift filtering, both theoretically and on a 2D orientation-data smoothing task. Channel smoothing is found to be superior in four respects: It has a lower computational complexity, it is easy to implement, it chooses the global minimum error instead of the nearest local minimum, and it can also be used on nonlinear spaces, such as orientation space. © 2006 IEEE.

  • 63.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Forssen, Per-Erik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Scharr, Hanno
    n/a.
    B-Spline Channel Smoothing for Robust Estimation2004Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we present a new method to implement a robust estimator: B-spline channel smoothing. We show that linear smoothing of channels is equivalent to a robust estimator, where we make use of the channel representation based upon quadratic B-splines. The linear decoding from B-spline channels allows to derive a robust error norm which is very similar to Tukey's biweight error norm. Using channel smoothing instead of iterative robust estimator implementations like non-linear diffusion, bilateral filtering, and mean-shift approaches is advantageous since channel smoothing is faster, it is easy to implement, it chooses the global minimum error instead of the nearest local minimum, and it can also be used on non-linear spaces, such as orientation space. As an application, we implemented orientation smoothing and compared it to the other three approaches.

  • 64.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Forssen, Per-Erik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Scharr, Hanno
    n/a.
    Efficient Robust Smoothing of Low-Level Signal Features2004Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we present a new and efficient method to implement robust smoothing of low-level signal features: B-spline channel smoothing. This method consists of three steps: encoding of the signal features into channels, averaging of the channels, and decoding of the channels. We show that linear smoothing of channels is equivalent to robust smoothing of the signal features, where we make use of quadratic B-splines to generate the channels. The linear decoding from B-spline channels allows to derive a robust error norm which is very similar to Tukey's biweight error norm. Channel smoothing is superior to iterative robust smoothing implementations like non-linear diffusion, bilateral filtering, and mean-shift approaches for four reasons: it has lower computational complexity, it is easy to implement, it chooses the global minimum error instead of the nearest local minimum, and it can also be used on non-linear spaces, such as orientation space. In the experimental part of the paper we compare channel smoothing and the previously mentioned three other approaches for 2D orientation data.

  • 65.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Forssén, Per-Erik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Moe, Anders
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Granlund, Gösta
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    A COSPAL Subsystem: Solving a Shape-Sorter Puzzle2005Ingår i: AAAI Fall Symposium: From Reactive to Anticipatory Cognitive Embedded Systems, FS-05-05, AAAI Press , 2005, s. 65-69Konferensbidrag (Refereegranskat)
    Abstract [en]

     To program a robot to solve a simple shape-sorter puzzle is trivial. To devise a Cognitive System Architecture, which allows the system to find out by itself how to go about a solution, is less than trivial. The development of such an architecture is one of the aims of the COSPAL project, leading to new techniques in vision based Artificial Cognitive Systems, which allow the development of robust systems for real dynamic environments. The systems developed under the project itself remain however in simplified scenarios, likewise the shape-sorter problem described in the present paper. The key property of the described system is its robustness. Since we apply association strategies of local features, the system behaves robustly under a wide range of distortions, as occlusion, colour and intensity changes. The segmentation step which is applied in many systems known from literature is replaced with local associations and view-based hypothesis validation. The hypotheses used in our system are based on the anticipated state of the visual percepts. This state replaces explicit modeling of shapes. The current state is chosen by a voting system and verified against the true visual percepts. The anticipated state is obtained from the association to the manipulator actions, where reinforcement learning replaces the explicit calculation of actions. These three differences to classical schemes allow the design of a much more generic and flexible system with a high level of robustness. On the technical side, the channel representation of information and associative learning in terms of the channel learning architecture are essential ingredients for the system. It is the properties of locality, smoothness, and non-negativity which make these techniques suitable for this kind of application. The paper gives brief descriptions of how different system parts have been implemented and show some examples from our tests.

  • 66.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Granlund, Gösta
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Anisotropic Channel Filtering2003Ingår i: SCIA: Gothenburg, Sweden, 2003, s. 755-762Konferensbidrag (Refereegranskat)
    Abstract [en]

    Channel smoothing is an alternative to diffusion filtering for robust estimation of image features. Its main advantages are speed, stability with respect to parameter changes, and a simple implementation. However, channel smoothing becomes instable in certain situations, typically for elongated, periodic patterns like for instance fingerprints. As for the diffusion filtering an anisotropic extension is required in these cases. In this paper we introduce a new method for anisotropic channel smoothing which is comparable to coherence enhancing diffusion, but faster and easier to implement. Anisotropic channel smoothing implements an orientation adaptive non-linear filtering scheme as a special case of adaptive channel filtering. The smoothing algorithm is applied to several fingerprint images and the results are compared to those of coherence enhancing diffusion.

  • 67.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Granlund, Gösta
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Fusing Dynamic Percepts and Symbols in Cognitive Systems2008Ingår i: International Conference on Cognitive Systems, 2008Konferensbidrag (Refereegranskat)
  • 68.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Granlund, Gösta
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    P-Channels: Robust Multivariate M-Estimation of Large Datasets2006Ingår i: ICPR,2006, 2006Konferensbidrag (Refereegranskat)
  • 69.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Granlund, Gösta
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    POI detection using channel clustering and the 2D energy tensor2004Ingår i: Proceedings of Pattern Recognition, 26th DAGM Symposium / [ed] Carl Edward Rasmussen, Heinrich H. Bülthoff, Bernhard Schölkopf and Martin A. Giese, SpringerLink , 2004, Vol. 3175, s. 103-110Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we address one of the standard problems of image processing and computer vision: The detection of points of interest (POI). We propose two new approaches for improving the detection results. First, we define an energy tensor which can be considered as a phase invariant extension of the structure tensor. Second, we use the channel representation for robustly clustering the POI information from the first step resulting in sub-pixel accuracy for the localisation of POI. We compare our method to several related approaches on a theoretical level and show a brief experimental comparison to the Harris detector.

  • 70.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Hedborg, Johan
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Real-Time View-Based Pose Recognition and Interpolation for Tracking Initialization2007Ingår i: Journal of Real-Time Image Processing, ISSN 1861-8200, E-ISSN 1861-8219, Journal of real-time image processing, ISSN 1861-8200, Vol. 2, nr 2-3, s. 103-115Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    In this paper we propose a new approach to real-time view-based pose recognition and interpolation. Pose recognition is particularly useful for identifying camera views in databases, video sequences, video streams, and live recordings. All of these applications require a fast pose recognition process, in many cases video real-time. It should further be possible to extend the database with new material, i.e., to update the recognition system online. The method that we propose is based on P-channels, a special kind of information representation which combines advantages of histograms and local linear models. Our approach is motivated by its similarity to information representation in biological systems but its main advantage is its robustness against common distortions such as clutter and occlusion. The recognition algorithm consists of three steps: (1) low-level image features for color and local orientation are extracted in each point of the image; (2) these features are encoded into P-channels by combining similar features within local image regions; (3) the query P-channels are compared to a set of prototype P-channels in a database using a least-squares approach. The algorithm is applied in two scene registration experiments with fisheye camera data, one for pose interpolation from synthetic images and one for finding the nearest view in a set of real images. The method compares favorable to SIFT-based methods, in particular concerning interpolation. The method can be used for initializing pose-tracking systems, either when starting the tracking or when the tracking has failed and the system needs to re-initialize. Due to its real-time performance, the method can also be embedded directly into the tracking system, allowing a sensor fusion unit choosing dynamically between the frame-by-frame tracking and the pose recognition.

  • 71.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Hedborg, Johan
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Real-Time Visual Recognition of Objects and Scenes Using P-Channel Matching2007Ingår i: Proceedings 15th Scandinavian Conference on Image Analysis / [ed] Bjarne K. Ersboll and Kim S. Pedersen, Berlin, Heidelberg: Springer, 2007, Vol. 4522, s. 908-917Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we propose a new approach to real-time view-based object recognition and scene registration. Object recognition is an important sub-task in many applications, as e.g., robotics, retrieval, and surveillance. Scene registration is particularly useful for identifying camera views in databases or video sequences. All of these applications require a fast recognition process and the possibility to extend the database with new material, i.e., to update the recognition system online. The method that we propose is based on P-channels, a special kind of information representation which combines advantages of histograms and local linear models. Our approach is motivated by its similarity to information representation in biological systems but its main advantage is its robustness against common distortions as clutter and occlusion. The recognition algorithm extracts a number of basic, intensity invariant image features, encodes them into P-channels, and compares the query P-channels to a set of prototype P-channels in a database. The algorithm is applied in a cross-validation experiment on the COIL database, resulting in nearly ideal ROC curves. Furthermore, results from scene registration with a fish-eye camera are presented.

  • 72.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Heyden, AndersLund University, Lund, Sweden.Krüger, NorbertUniversity of Southern Denmark, Odense, Denmark.
    Computer Analysis of Images and Patterns: 17th International Conference, CAIP 2017, Ystad, Sweden, August 22-24, 2017, Proceedings, Part I2017Proceedings (redaktörskap) (Refereegranskat)
    Abstract [en]

    The two volume set LNCS 10424 and 10425 constitutes the refereed proceedings of the 17th International Conference on Computer Analysis of Images and Patterns, CAIP 2017, held in Ystad, Sweden, in August 2017.

    The 72 papers presented were carefully reviewed and selected from 144 submissions The papers are organized in the following topical sections: Vision for Robotics; Motion and Tracking; Segmentation; Image/Video Indexing and Retrieval; Shape Representation and Analysis; Biomedical Image Analysis; Biometrics; Machine Learning; Image Restoration; and Poster Sessions.

  • 73.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Heyden, AndersLund University, Lund, Sweden.Krüger, NorbertUniversity of Southern Denmark, Odense, Denmark.
    Computer Analysis of Images and Patterns: 17th International Conference, CAIP 2017, Ystad, Sweden, August 22-24, 2017, Proceedings, Part II2017Proceedings (redaktörskap) (Refereegranskat)
    Abstract [en]

    The two volume set LNCS 10424 and 10425 constitutes the refereed proceedings of the 17th International Conference on Computer Analysis of Images and Patterns, CAIP 2017, held in Ystad, Sweden, in August 2017.  The 72 papers presented were carefully reviewed and selected from 144 submissions The papers are organized in the following topical sections: Vision for Robotics; Motion and Tracking; Segmentation; Image/Video Indexing and Retrieval; Shape Representation and Analysis; Biomedical Image Analysis; Biometrics; Machine Learning; Image Restoration; and Poster Sessions.

  • 74.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Jonsson, Erik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Energy Tensors: Quadratic, Phase Invariant Image Operators2005Ingår i: Pattern Recognition: 27th DAGM Symposium, Vienna, Austria, August 31 - September 2, 2005. Proceedings / [ed] Walter G. Kropatsch, Robert Sablatnig and Allan Hanbury, Springer Berlin/Heidelberg, 2005, Vol. 3663, s. 493-500Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we briefly review a not so well known quadratic, phase invariant image processing operator, the energy operator, and describe its tensor-valued generalization, the energy tensor. We present relations to the real-valued and the complex valued energy operators and discuss properties of the three operators. We then focus on the discrete implementation for estimating the tensor based on Teager’s algorithm and frame theory. The kernels of the real-valued and the tensor-valued operators are formally derived. In a simple experiment we compare the energy tensor to other operators for orientation estimation. The paper is concluded with a short outlook to future work.

  • 75.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Jonsson, Erik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Reconstruction of Probability Density Functions from Channel Representations2005Ingår i: Reconstruction of Probability Density Functions from Channel Representations,2005, 2005Konferensbidrag (Refereegranskat)
    Abstract [en]

    The channel representation allows the construction of soft histograms, where peaks can be detected with a much higher accuracy than in regular hard-binned histograms. This is critical in e.g. reducing the number of bins of generalized Hough transform methods. When applying the maximum entropy method to the channel representation, a minimum-information reconstruction of the underlying continuous probability distribution is obtained. The maximum entropy reconstruction is compared to simpler linear methods in some simulated situations. Experimental results show that mode estimation of the maximum entropy reconstruction outperforms the linear methods in terms of quantization error and discrimination threshold. Finding the maximum entropy reconstruction is however computationally more expensive.

  • 76.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Koch, Reinhard
    Editorial for the special issue on markerless real-time tracking for augmented reality image synthesis2007Ingår i: Journal of Real-Time Image Processing, ISSN 1861-8200, E-ISSN 1861-8219, Vol. 2, nr 2-3, s. 67-68Artikel i tidskrift (Övrigt vetenskapligt)
    Abstract [en]

    Augmented reality is a growing field, with many diverse applications ranging from TV and film production, to industrial maintenance, medicine, education, entertainment and games. The central idea is to add virtual objects into a real scene, either by displaying them in a see-through head-mounted display, or by superimposing them on an image of the scene captured by a camera. Depending on the application, the added objects might be virtual characters in a TV or film production, instructions for repairing a car engine, or a reconstruction of an archaeological site. For the effect to be believable, the virtual objects must appear rigidly fixed to the real world, which requires the accurate measurement in real-time of the position of the camera or the user-s head. Present technology cannot achieve this without resorting to systems that require a significant infrastructure in the operating environment, severely restricting the range of possible applications.

  • 77.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten. Linköpings universitet, Centrum för medicinsk bildvetenskap och visualisering, CMIV.
    Kristan, Matej
    University of Ljubljana, Slovenia.
    Matas, Jiri
    Czech Technical University, Czech Republic.
    Leonardis, Ales
    University of Birmingham, England.
    Pflugfelder, Roman
    Austrian Institute Technology, Austria.
    Häger, Gustav
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Berg, Amanda
    Linköpings universitet, Tekniska fakulteten. Linköpings universitet, Institutionen för systemteknik, Datorseende. Termisk Syst Tekn AB, Linkoping, Sweden.
    Eldesokey, Abdelrahman
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Ahlberg, Jörgen
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten. Termisk Syst Tekn AB, Linkoping, Sweden.
    Cehovin, Luka
    University of Ljubljana, Slovenia.
    Vojir, Tomas
    Czech Technical University, Czech Republic.
    Lukezic, Alan
    University of Ljubljana, Slovenia.
    Fernandez, Gustavo
    Austrian Institute Technology, Austria.
    Petrosino, Alfredo
    Parthenope University of Naples, Italy.
    Garcia-Martin, Alvaro
    University of Autonoma Madrid, Spain.
    Solis Montero, Andres
    University of Ottawa, Canada.
    Varfolomieiev, Anton
    Kyiv Polytech Institute, Ukraine.
    Erdem, Aykut
    Hacettepe University, Turkey.
    Han, Bohyung
    POSTECH, South Korea.
    Chang, Chang-Ming
    University of Albany, GA USA.
    Du, Dawei
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Erdem, Erkut
    Hacettepe University, Turkey.
    Khan, Fahad Shahbaz
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Porikli, Fatih
    ARC Centre Excellence Robot Vis, Australia; CSIRO, Australia.
    Zhao, Fei
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Bunyak, Filiz
    University of Missouri, MO 65211 USA.
    Battistone, Francesco
    Parthenope University of Naples, Italy.
    Zhu, Gao
    University of Missouri, Columbia, USA.
    Seetharaman, Guna
    US Navy, DC 20375 USA.
    Li, Hongdong
    ARC Centre Excellence Robot Vis, Australia.
    Qi, Honggang
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Bischof, Horst
    Graz University of Technology, Austria.
    Possegger, Horst
    Graz University of Technology, Austria.
    Nam, Hyeonseob
    NAVER Corp, South Korea.
    Valmadre, Jack
    University of Oxford, England.
    Zhu, Jianke
    Zhejiang University, Peoples R China.
    Feng, Jiayi
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Lang, Jochen
    University of Ottawa, Canada.
    Martinez, Jose M.
    University of Autonoma Madrid, Spain.
    Palaniappan, Kannappan
    University of Missouri, MO 65211 USA.
    Lebeda, Karel
    University of Surrey, England.
    Gao, Ke
    University of Missouri, MO 65211 USA.
    Mikolajczyk, Krystian
    Imperial Coll London, England.
    Wen, Longyin
    University of Albany, GA USA.
    Bertinetto, Luca
    University of Oxford, England.
    Poostchi, Mahdieh
    University of Missouri, MO 65211 USA.
    Maresca, Mario
    Parthenope University of Naples, Italy.
    Danelljan, Martin
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Arens, Michael
    Fraunhofer IOSB, Germany.
    Tang, Ming
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Baek, Mooyeol
    POSTECH, South Korea.
    Fan, Nana
    Harbin Institute Technology, Peoples R China.
    Al-Shakarji, Noor
    University of Missouri, MO 65211 USA.
    Miksik, Ondrej
    University of Oxford, England.
    Akin, Osman
    Hacettepe University, Turkey.
    Torr, Philip H. S.
    University of Oxford, England.
    Huang, Qingming
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Martin-Nieto, Rafael
    University of Autonoma Madrid, Spain.
    Pelapur, Rengarajan
    University of Missouri, MO 65211 USA.
    Bowden, Richard
    University of Surrey, England.
    Laganiere, Robert
    University of Ottawa, Canada.
    Krah, Sebastian B.
    Fraunhofer IOSB, Germany.
    Li, Shengkun
    University of Albany, GA USA.
    Yao, Shizeng
    University of Missouri, MO 65211 USA.
    Hadfield, Simon
    University of Surrey, England.
    Lyu, Siwei
    University of Albany, GA USA.
    Becker, Stefan
    Fraunhofer IOSB, Germany.
    Golodetz, Stuart
    University of Oxford, England.
    Hu, Tao
    Australian National University, Australia; Chinese Academic Science, Peoples R China.
    Mauthner, Thomas
    Graz University of Technology, Austria.
    Santopietro, Vincenzo
    Parthenope University of Naples, Italy.
    Li, Wenbo
    Lehigh University, PA 18015 USA.
    Huebner, Wolfgang
    Fraunhofer IOSB, Germany.
    Li, Xin
    Harbin Institute Technology, Peoples R China.
    Li, Yang
    Zhejiang University, Peoples R China.
    Xu, Zhan
    Zhejiang University, Peoples R China.
    He, Zhenyu
    Harbin Institute Technology, Peoples R China.
    The Thermal Infrared Visual Object Tracking VOT-TIR2016 Challenge Results2016Ingår i: Computer Vision – ECCV 2016 Workshops. ECCV 2016. / [ed] Hua G., Jégou H., SPRINGER INT PUBLISHING AG , 2016, s. 824-849Konferensbidrag (Refereegranskat)
    Abstract [en]

    The Thermal Infrared Visual Object Tracking challenge 2016, VOT-TIR2016, aims at comparing short-term single-object visual trackers that work on thermal infrared (TIR) sequences and do not apply pre-learned models of object appearance. VOT-TIR2016 is the second benchmark on short-term tracking in TIR sequences. Results of 24 trackers are presented. For each participating tracker, a short description is provided in the appendix. The VOT-TIR2016 challenge is similar to the 2015 challenge, the main difference is the introduction of new, more difficult sequences into the dataset. Furthermore, VOT-TIR2016 evaluation adopted the improvements regarding overlap calculation in VOT2016. Compared to VOT-TIR2015, a significant general improvement of results has been observed, which partly compensate for the more difficult sequences. The dataset, the evaluation kit, as well as the results are publicly available at the challenge website.

  • 78.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Kruger, Norbert
    n/a.
    A Probabilistic Definition of Intrinsic Dimensionality for Images2003Ingår i: 25. DAGM Symposium Mustererkennung, Magdeburg eds Michaelis, B. and Krell, G., 2003, Vol. 2781, s. 140-147Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we address the problem of appropriately representing the intrinsic dimensionality of image neighborhoods. This dimensionality describes the degrees of freedom of a local image patch and it gives rise to some of the most often applied corner and edge detectors. It is common to categorize the intrinsic dimensionality (iD) to three distinct cases: i0D, i1D, and i2D. Real images however contain combinations of all three dimensionalities which has to be taken into account by a continuous representation. Based on considerations of the structure tensor, we derive a cone-shaped iD-space which leads to a probabilistic point of view to the estimation of intrinsic dimensionality.

  • 79.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Kruger, Norbert
    n/a.
    A Probabilistic Definition of Intrinsic Dimensionality for Images2003Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we address the problem of appropriately representing the intrinsic dimensionality of image neighborhoods. This dimensionality describes the degrees of freedom of a local image patch and it gives rise to some of the most often applied corner and edge detectors. It is common to categorize the intrinsic dimensionality (iD) to three distinct cases: i0D, i1D, and i2D. Real images however contain combinations of all three dimensionalities which has to be taken into account by a continuous representation. Based on considerations of the structure tensor, we derive a cone-shaped iD-space which leads to a probabilistic point of view to the estimation of intrinsic dimensionality.

  • 80.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Köthe, Ullrich
    GET: The connection between monogenic scale-space and Gaussian derivatives2005Ingår i: Scale Space and PDE Methods in Computer Vision, 2005, Vol. 3459, s. 192-203Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we propose a new operator which combines advantages of monogenic scale-space and Gaussian scale-space, of the monogenic signal and the structure tensor. The gradient energy tensor (GET) defined in this paper is based on Gaussian derivatives up to third order using different scales. These filters are commonly available, separable, and have an optimal uncertainty. The response of this new operator can be used like the monogenic signal to estimate the local amplitude, the local phase, and the local orientation of an image, but it also allows to measure the coherence of image regions as in the case of the structure tensor. Both theoretically and in experiments the new approach compares favourably with existing methods.

  • 81.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Larsson, Fredrik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Learning Bayesian tracking for motion estimation2008Ingår i: ECCV Workshop: Machine Learning for Vision-based Motion Analysis, 2008Konferensbidrag (Refereegranskat)
  • 82.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Larsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Learning Higher-Order Markov Models for ObjectTracking in Image Sequences2009Ingår i: Proceedings of the 5th International Symposium on Advances in Visual Computing: Part II, Berlin, Heidelberg: Springer-Verlag , 2009, s. 184-195Konferensbidrag (Refereegranskat)
    Abstract [en]

    This work presents a novel object tracking approach, where the motion model is learned from sets of frame-wise detections with unknown associations. We employ a higher-order Markov model on position space instead of a first-order Markov model on a high-dimensional state-space of object dynamics. Compared to the latter, our approach allows the use of marginal rather than joint distributions, which results in a significant reduction of computation complexity. Densities are represented using a grid-based approach, where the rectangular windows are replaced with estimated smooth Parzen windows sampled at the grid points. This method performs as accurately as particle filter methods with the additional advantage that the prediction and update steps can be learned from empirical data. Our method is compared against standard techniques on image sequences obtained from an RC car following scenario. We show that our approach performs best in most of the sequences. Other potential applications are surveillance from cheap or uncalibrated cameras and image sequence analysis.

  • 83.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Larsson, Fredrik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Learning object tracking in image sequences2010Ingår i: International Conference on Cognitive Systems, 2010Konferensbidrag (Övrigt vetenskapligt)
  • 84.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Larsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Han, Wang
    Nanyang Technological University, China.
    Ynnerman, Anders
    Linköpings universitet, Institutionen för teknik och naturvetenskap, Medie- och Informationsteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Torch Guided Navigation2010Ingår i: Proceedings of the 2010 SSBA Symposium, 2010, s. 8-9Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    A common computer vision task is navigation and mapping. Many indoor navigation tasks require depth knowledge of at, unstructured surfaces (walls, oor, ceiling). With passive illumination only, this is an ill-posed problem. Inspired by small children using a torchlight, we use a spotlight for active illumination. Using our torchlight approach, depth and orientation estimation of unstructured, at surfaces boils down to estimation of ellipse parameters. The extraction of ellipses is very robust and requires little computational effort.

  • 85.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Larsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Wang, Han
    National University Singapore.
    Ynnerman, Anders
    Linköpings universitet, Centrum för medicinsk bildvetenskap och visualisering, CMIV. Linköpings universitet, Institutionen för teknik och naturvetenskap, Medie- och Informationsteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Torchlight Navigation2010Ingår i: Proceedings of the 20th International Conferenceon Pattern Recognition, 2010, s. 302-306Konferensbidrag (Refereegranskat)
    Abstract [en]

    A common computer vision task is navigation and mapping. Many indoor navigation tasks require depth knowledge of flat, unstructured surfaces (walls, floor, ceiling). With passive illumination only, this is an ill-posed problem. Inspired by small children using a torchlight, we use a spotlight for active illumination. Using our torchlight approach, depth and orientation estimation of unstructured, flat surfaces boils down to estimation of ellipse parameters. The extraction of ellipses is very robust and requires little computational effort.

  • 86.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Larsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska högskolan.
    Wiklund, Johan
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska högskolan.
    Wadströmer, Niclas
    FOI.
    Ahlberg, Jörgen
    Termisk Systemteknik AB.
    Online Learning of Correspondences between Images2013Ingår i: IEEE Transaction on Pattern Analysis and Machine Intelligence, ISSN 0162-8828, E-ISSN 1939-3539, Vol. 35, nr 1, s. 118-129Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    We propose a novel method for iterative learning of point correspondences between image sequences. Points moving on surfaces in 3D space are projected into two images. Given a point in either view, the considered problem is to determine the corresponding location in the other view. The geometry and distortions of the projections are unknown as is the shape of the surface. Given several pairs of point-sets but no access to the 3D scene, correspondence mappings can be found by excessive global optimization or by the fundamental matrix if a perspective projective model is assumed. However, an iterative solution on sequences of point-set pairs with general imaging geometry is preferable. We derive such a method that optimizes the mapping based on Neyman's chi-square divergence between the densities representing the uncertainties of the estimated and the actual locations. The densities are represented as channel vectors computed with a basis function approach. The mapping between these vectors is updated with each new pair of images such that fast convergence and high accuracy are achieved. The resulting algorithm runs in real-time and is superior to state-of-the-art methods in terms of convergence and accuracy in a number of experiments.

  • 87.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Scharr, Hanno
    n/a.
    Forssen, Per-Erik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    The B-Spline Channel Representation: Channel Algebra and Channel Based Diffusion Filtering2002Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we consider the channel representation based upon quadratic B-splines from a statistical point of view. Interpreting the channel representation as a kernel method for estimating probability density functions, we establish a channel algebra which allows to perform basic algebraic operations on measurements directly in the channel representation. Furthermore, as a central point, we identify the smoothing of channel values with a robust estimator, or equivalently, a diffusion process.

  • 88.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik.
    Scharr, Hanno
    Forssén, Per-Erik
    Linköpings universitet, Institutionen för systemteknik.
    B-spline channel smoothing for robust estimation 2004Rapport (Övrigt vetenskapligt)
  • 89.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Shaukat, Affan
    University of Surrey, Guildford, UK..
    Windridge, David
    University of Surrey, Guildford, UK..
    Online Learning in Perception-Action Systems2010Ingår i: ECCV 2010 Workshop on Vision for Cognitive Tasks, 2010Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this position paper, we seek to extend the layered perception-action paradigm for on-line learning such that it includes an explicit symbolic processing capability. By incorporating symbolic processing at the apex of the perception action hierarchy in this way, we ensure that abstract symbol manipulation is fully grounded, without the necessity of specifying an explicit representational framework. In order to carry out this novel interfacing between symbolic and sub-symbolic processing, it is necessary to embed fuzzy rst-order logic theorem proving within a variational framework. The online learning resulting from the corresponding Euler-Lagrange equations establishes an extended adaptability compared to the standard subsumption architecture. We discuss an application of this approach within the eld of advanced driver assistance systems, demonstrating that a closed-form solution to the Euler Lagrange optimization problem is obtainable for simple cases.

     

  • 90.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Sommer, Gerald
    n/a.
    Image Features Based on a New Approach to 2D Rotation Invariant Quadrature Filters2002Ingår i: Computer Vision - ECCV 2002 eds A. Heyden and G. Sparr and M. Nielsen and P. Johansen, 2002, Vol. 2350, s. 369-383Konferensbidrag (Refereegranskat)
    Abstract [en]

    Quadrature filters are a well known method of low-level computer vision for estimating certain properties of the signal, as there are local amplitude and local phase. However, 2D quadrature filters suffer from being not rotation invariant. Furthermore, they do not allow to detect truly 2D features as corners and junctions unless they are combined to form the structure tensor. The present paper deals with a new 2D generalization of quadrature filters which is rotation invariant and allows to analyze intrinsically 2D signals. Hence, the new approach can be considered as the union of properties of quadrature filters and of the structure tensor. The proposed method first estimates the local orientation of the signal which is then used for steering some basis filter responses. Certain linear combination of these filter responses are derived which allow to estimate the local isotropy and two perpendicular phases of the signal. The phase model is based on the assumption of an angular band-limitation in the signal. As an application, a simple and efficient point-of-interest operator is presented and it is compared to the Plessey detector.

  • 91.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Sommer, Gerald
    The Monogenic Scale-Space: A Unifying Approach to Phase-Based Image Processing in Scale-Space2004Ingår i: Journal of Mathematical Imaging and Vision, ISSN 0924-9907, E-ISSN 1573-7683, Vol. 21Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    In this paper we address the topics of scale-space and phase-based image processing in a unifying framework. In contrast to the common opinion, the Gaussian kernel is not the unique choice for a linear scale-space. Instead, we chose the Poisson kernel since it is closely related to the monogenic signal, a 2D generalization of the analytic signal, where the Riesz transform replaces the Hilbert transform. The Riesz transform itself yields the flux of the Poisson scale-space and the combination of flux and scale-space, the monogenic scale-space, provides the local features phase-vector and attenuation in scale-space. Under certain assumptions, the latter two again form a monogenic scale-space which gives deeper insight to low-level image processing. In particular, we discuss edge detection by a new approach to phase congruency and its relation to amplitude based methods, reconstruction from local amplitude and local phase, and the evaluation of the local frequency.

     

  • 92.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Sommer, Gerald
    n/a.
    The monogenic signal2001Ingår i: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 49, nr 12, s. 3136-3144Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    This paper introduces a two-dimensional generalization of the analytic signal. This novel approach is based on the Riesz transform, which is used instead of the Hilbert transform. The combination of a 2D signal with the Riesz transformed one yields a sophisticated 2D analytic signal, the monogenic signal. The approach is derived analytically from irrotational and solenoidal vector fields. Based on local amplitude and local phase, an appropriate local signal representation is presented which preserves the split of identity, i.e., the invariance – equivariance property of signal decomposition. This is one of the central properties of the 1D analytic signal that decomposes a signal into structural and energetic information. We show that further properties of the analytic signal concerning symmetry, energy, allpass transfer function, and orthogonality are also preserved, and we compare this to the behavior of other approaches for a 2D analytic signal. As a central topic of this paper, a geometric phase interpretation is introduced which is based on the relation between the 1D analytic signal and the 2D monogenic signal established by the Radon transform. Possible applications of this relationship are sketched and references to other applications of the monogenic signal are given. This report is a revised version of the technical report 2009 [7], and therefore supercedes it.

  • 93.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Sommer, Gerald
    n/a.
    The Poisson Scale-Space: A Unified Approach to Phase-Based Image Processing in Scale-Space2002Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we address the topics of scale-space and phase-based signal processing in a common framework. The involved linear scale-space is no longer based on the Gaussian kernel but on the Poisson kernel. The resulting scale-space representation is directly related to the monogenic signal, a 2D generalization of the analytic signal. Hence, the local phase arises as a natural concept in this framework which results in several advanced relationships that can be used in image processing.

  • 94.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Wiklund, Johan
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Granlund, Gösta
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Exploratory learning structures in artificial cognitive systems2009Ingår i: Image and Vision Computing, ISSN 0262-8856, Vol. 27, nr 11, s. 1671-1687Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    The major goal of the COSPAL project is to develop an artificial cognitive system architecture, with the ability to autonomously extend its capabilities. Exploratory learning is one strategy that allows an extension of competences as provided by the environment of the system. Whereas classical learning methods aim at best for a parametric generalization, i.e., concluding from a number of samples of a problem class to the problem class itself, exploration aims at applying acquired competences to a new problem class, and to apply generalization on a conceptual level, resulting in new models. Incremental or online learning is a crucial requirement to perform exploratory learning. In the COSPAL project, we mainly investigate reinforcement-type learning methods for exploratory learning, and in this paper we focus on the organization of cognitive systems for efficient operation. Learning is used over the entire system. It is organized in the form of four nested loops, where the outermost loop reflects the user-reinforcement-feedback loop, the intermediate two loops switch between different solution modes at symbolic respectively sub-symbolic level, and the innermost loop performs the acquired competences in terms of perception-action cycles. We present a system diagram which explains this process in more detail. We discuss the learning strategy in terms of learning scenarios provided by the user. This interaction between user (teacher) and system is a major difference to classical robotics systems, where the system designer places his world model into the system. We believe that this is the key to extendable robust system behavior and successful interaction of humans and artificial cognitive systems. We furthermore address the issue of bootstrapping the system, and, in particular, the visual recognition module. We give some more in-depth details about our recognition method and how feedback from higher levels is implemented. The described system is however work in progress and no final results are available yet. The available preliminary results that we have achieved so far, clearly point towards a successful proof of the architecture concept.

  • 95.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Wiklund, Johan
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Jonsson, Erik
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Moe, Anders
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Granlund, Gösta
    Linköpings universitet, Institutionen för systemteknik, Bildbehandling. Linköpings universitet, Tekniska högskolan.
    Exploratory Learning Structure in Artificial Cognitive Systems2006Rapport (Övrigt vetenskapligt)
    Abstract [en]

    One major goal of the COSPAL project is to develop an artificial cognitive system architecture with the capability of exploratory learning. Exploratory learning is a strategy that allows to apply generalization on a conceptual level, resulting in an extension of competences. Whereas classical learning methods aim at best possible generalization, i.e., concluding from a number of samples of a problem class to the problem class itself, exploration aims at applying acquired competences to a new problem class. Incremental or online learning is an inherent requirement to perform exploratory learning.

    Exploratory learning requires new theoretic tools and new algorithms. In the COSPAL project, we mainly investigate reinforcement-type learning methods for exploratory learning and in this paper we focus on its algorithmic aspect. Learning is performed in terms of four nested loops, where the outermost loop reflects the user-reinforcement-feedback loop, the intermediate two loops switch between different solution modes at symbolic respectively sub-symbolic level, and the innermost loop performs the acquired competences in terms of perception-action cycles. We present a system diagram which explains this process in more detail.

    We discuss the learning strategy in terms of learning scenarios provided by the user. This interaction between user (’teacher’) and system is a major difference to most existing systems where the system designer places his world model into the system. We believe that this is the key to extendable robust system behavior and successful interaction of humans and artificial cognitive systems.

    We furthermore address the issue of bootstrapping the system, and, in particular, the visual recognition module.We give some more in-depth details about our recognition method and how feedback from higher levels is implemented. The described system is however work in progress and no final results are available yet. The available preliminary results that we have achieved so far, clearly point towards a successful proof of the architecture concept.

  • 96.
    Felsberg, Michael
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Wiklund, Johan
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Jonsson, Erik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Moe, Anders
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Granlund, Gösta
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Bildbehandling.
    Exploratory Learning Strucutre in Artificial Cognitive Systems2007Ingår i: International Cognitive Vision Workshop, Bielefeld: eCollections , 2007Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    One major goal of the COSPAL project is to develop an artificial cognitive system architecture with the capability of exploratory learning. Exploratory learning is a strategy that allows to apply generalization on a conceptual level, resulting in an extension of competences. Whereas classical learning methods aim at best possible generalization, i.e., concluding from a number of samples of a problem class to the problem class itself, exploration aims at applying acquired competences to a new problem class. Incremental or online learning is an inherent requirement to perform exploratory learning.

    Exploratory learning requires new theoretic tools and new algorithms. In the COSPAL project, we mainly investigate reinforcement-type learning methods for exploratory learning and in this paper we focus on its algorithmic aspect. Learning is performed in terms of four nested loops, where the outermost loop reflects the user-reinforcement-feedback loop, the intermediate two loops switch between different solution modes at symbolic respectively sub-symbolic level, and the innermost loop performs the acquired competences in terms of perception-action cycles. We present a system diagram which explains this process in more detail.

    We discuss the learning strategy in terms of learning scenarios provided by the user. This interaction between user ('teacher') and system is a major difference to most existing systems where the system designer places his world model into the system. We believe that this is the key to extendable robust system behavior and successful interaction of humans and artificial cognitive systems.

    We furthermore address the issue of bootstrapping the system, and, in particular, the visual recognition module. We give some more in-depth details about our recognition method and how feedback from higher levels is implemented. The described system is however work in progress and no final results are available yet. The available preliminary results that we have achieved so far, clearly point towards a successful proof of the architecture concept.

  • 97.
    Felsberg, Michael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Öfjäll, Kristoffer
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Lenz, Reiner
    Linköpings universitet, Institutionen för teknik och naturvetenskap, Medie- och Informationsteknik. Linköpings universitet, Tekniska fakulteten.
    Unbiased decoding of biologically motivated visual feature descriptors2015Ingår i: Frontiers in Robotics and AI, ISSN 2296-9144, Vol. 2, nr 20Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    Visual feature descriptors are essential elements in most computer and robot vision systems. They typically lead to an abstraction of the input data, images, or video, for further processing, such as clustering and machine learning. In clustering applications, the cluster center represents the prototypical descriptor of the cluster and estimates the corresponding signal value, such as color value or dominating flow orientation, by decoding the prototypical descriptor. Machine learning applications determine the relevance of respective descriptors and a visualization of the corresponding decoded information is very useful for the analysis of the learning algorithm. Thus decoding of feature descriptors is a relevant problem, frequently addressed in recent work. Also, the human brain represents sensorimotor information at a suitable abstraction level through varying activation of neuron populations. In previous work, computational models have been derived that agree with findings of neurophysiological experiments on the represen-tation of visual information by decoding the underlying signals. However, the represented variables have a bias toward centers or boundaries of the tuning curves. Despite the fact that feature descriptors in computer vision are motivated from neuroscience, the respec-tive decoding methods have been derived largely independent. From first principles, we derive unbiased decoding schemes for biologically motivated feature descriptors with a minimum amount of redundancy and suitable invariance properties. These descriptors establish a non-parametric density estimation of the underlying stochastic process with a particular algebraic structure. Based on the resulting algebraic constraints, we show formally how the decoding problem is formulated as an unbiased maximum likelihood estimator and we derive a recurrent inverse diffusion scheme to infer the dominating mode of the distribution. These methods are evaluated in experiments, where stationary points and bias from noisy image data are compared to existing methods.

  • 98.
    Freddie, Åström
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska högskolan.
    Michael, Felsberg
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska högskolan.
    Reiner, Lenz
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Institutionen för teknik och naturvetenskap, Medie- och Informationsteknik. Linköpings universitet, Tekniska högskolan.
    Color Persistent Anisotropic Diffusion of Images2011Ingår i: Image Analysis / [ed] Anders Heyden, Fredrik Kahl, Heidelberg: Springer, 2011, s. 262-272Konferensbidrag (Refereegranskat)
    Abstract [en]

    Techniques from the theory of partial differential equations are often used to design filter methods that are locally adapted to the image structure. These techniques are usually used in the investigation of gray-value images. The extension to color images is non-trivial, where the choice of an appropriate color space is crucial. The RGB color space is often used although it is known that the space of human color perception is best described in terms of non-euclidean geometry, which is fundamentally different from the structure of the RGB space. Instead of the standard RGB space, we use a simple color transformation based on the theory of finite groups. It is shown that this transformation reduces the color artifacts originating from the diffusion processes on RGB images. The developed algorithm is evaluated on a set of real-world images, and it is shown that our approach exhibits fewer color artifacts compared to state-of-the-art techniques. Also, our approach preserves details in the image for a larger number of iterations.

  • 99.
    Gladh, Susanna
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Datorseende.
    Danelljan, Martin
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Datorseende.
    Khan, Fahad Shahbaz
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Datorseende.
    Felsberg, Michael
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Datorseende.
    Deep motion features for visual tracking2016Ingår i: Proceedings of the 23rd International Conference on, Pattern Recognition (ICPR), 2016, Institute of Electrical and Electronics Engineers (IEEE), 2016, s. 1243-1248Konferensbidrag (Refereegranskat)
    Abstract [en]

    Robust visual tracking is a challenging computer vision problem, with many real-world applications. Most existing approaches employ hand-crafted appearance features, such as HOG or Color Names. Recently, deep RGB features extracted from convolutional neural networks have been successfully applied for tracking. Despite their success, these features only capture appearance information. On the other hand, motion cues provide discriminative and complementary information that can improve tracking performance. Contrary to visual tracking, deep motion features have been successfully applied for action recognition and video classification tasks. Typically, the motion features are learned by training a CNN on optical flow images extracted from large amounts of labeled videos. This paper presents an investigation of the impact of deep motion features in a tracking-by-detection framework. We further show that hand-crafted, deep RGB, and deep motion features contain complementary information. To the best of our knowledge, we are the first to propose fusing appearance information with deep motion features for visual tracking. Comprehensive experiments clearly suggest that our fusion approach with deep motion features outperforms standard methods relying on appearance information alone.

  • 100.
    Grelsson, Bertil
    et al.
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Felsberg, Michael
    Linköpings universitet, Institutionen för systemteknik, Datorseende. Linköpings universitet, Tekniska fakulteten.
    Improved Learning in Convolutional Neural Networks with Shifted Exponential Linear Units (ShELUs)2018Ingår i: 2018 24th International Conference on Pattern Recognition (ICPR), IEEE, 2018, s. 517-522Konferensbidrag (Refereegranskat)
    Abstract [en]

    The Exponential Linear Unit (ELU) has been proven to speed up learning and improve the classification performance over activation functions such as ReLU and Leaky ReLU for convolutional neural networks. The reasons behind the improved behavior are that ELU reduces the bias shift, it saturates for large negative inputs and it is continuously differentiable. However, it remains open whether ELU has the optimal shape and we address the quest for a superior activation function.We use a new formulation to tune a piecewise linear activation function during training, to investigate the above question, and learn the shape of the locally optimal activation function. With this tuned activation function, the classification performance is improved and the resulting, learned activation function shows to be ELU-shaped irrespective if it is initialized as a RELU, LReLU or ELU. Interestingly, the learned activation function does not exactly pass through the origin indicating that a shifted ELU-shaped activation function is preferable. This observation leads us to introduce the Shifted Exponential Linear Unit (ShELU) as a new activation function.Experiments on Cifar-100 show that the classification performance is further improved when using the ShELU activation function in comparison with ELU. The improvement is achieved when learning an individual bias shift for each neuron.

1234 51 - 100 av 195
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf