Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008105
Shampa Shahriyar, M. Murshed, Mortuza Ali, M. Paul
With the growing demands for 3D and multi-view video content, efficient depth data coding becomes a vital issue in image and video coding area. In this paper, we propose a simple depth coding scheme using multiple prediction modes exploiting temporal correlation of depth map. Current depth coding techniques mostly depend on intra-coding mode that cannot get the advantage of temporal redundancy in the depth maps and higher spatial redundancy in inter-predicted depth residuals. Depth maps are characterized by smooth regions with sharp edges that play an important role in the view synthesis process. As depth maps are more sensitive to coding errors, use of transformation or approximation of edges by explicit edge modelling has impact on view synthesis quality. Moreover, lossy compression of depth map brings additional geometrical distortion to synthetic view. In this paper, we have demonstrated that encoding inter-coded depth block residuals with quantization at pixel domain is more efficient than the intra-coding techniques relying on explicit edge preservation. On standard 3D video sequences, the proposed depth coding has achieved superior image quality of synthesized views against the new 3D-HEVC standard for depth map bit-rate 0.25 bpp or higher.
{"title":"Efficient Coding of Depth Map by Exploiting Temporal Correlation","authors":"Shampa Shahriyar, M. Murshed, Mortuza Ali, M. Paul","doi":"10.1109/DICTA.2014.7008105","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008105","url":null,"abstract":"With the growing demands for 3D and multi-view video content, efficient depth data coding becomes a vital issue in image and video coding area. In this paper, we propose a simple depth coding scheme using multiple prediction modes exploiting temporal correlation of depth map. Current depth coding techniques mostly depend on intra-coding mode that cannot get the advantage of temporal redundancy in the depth maps and higher spatial redundancy in inter-predicted depth residuals. Depth maps are characterized by smooth regions with sharp edges that play an important role in the view synthesis process. As depth maps are more sensitive to coding errors, use of transformation or approximation of edges by explicit edge modelling has impact on view synthesis quality. Moreover, lossy compression of depth map brings additional geometrical distortion to synthetic view. In this paper, we have demonstrated that encoding inter-coded depth block residuals with quantization at pixel domain is more efficient than the intra-coding techniques relying on explicit edge preservation. On standard 3D video sequences, the proposed depth coding has achieved superior image quality of synthesized views against the new 3D-HEVC standard for depth map bit-rate 0.25 bpp or higher.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116549790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008095
Nabeel Mohammed, D. Squire
In this paper we evaluate the utility of sparseness as a criterion for selecting a sub-set of independent component filters (ICF). Four sparseness measures were presented more than a decade ago by Le Borgne et al., but have since been ignored for ICF selection. In this paper we present our evaluation in the context of texture retrieval. We compare the sparseness-based method with the dispersal-based method, also proposed by Le Borgne et al., and the clustering-based method previously proposed by us. We show that the sparse filters and highly dispersed filters are quite different. In fact we show that highly dispersed filters tend to have lower sparseness. We also show that the sparse filters give better results compared to the highly dispersed filters when applied to texture retrieval. However the sparseness measures are calculated over filter response energies, making this method susceptible to choosing a redundant filter set. This issue is demonstrated and we show that ICF selected using our clustering-based method, which chooses a filter set with much lower redundancy, outperforms the sparse filters.
{"title":"An Evaluation of Sparseness as a Criterion for Selecting Independent Component Filters, When Applied to Texture Retrieval","authors":"Nabeel Mohammed, D. Squire","doi":"10.1109/DICTA.2014.7008095","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008095","url":null,"abstract":"In this paper we evaluate the utility of sparseness as a criterion for selecting a sub-set of independent component filters (ICF). Four sparseness measures were presented more than a decade ago by Le Borgne et al., but have since been ignored for ICF selection. In this paper we present our evaluation in the context of texture retrieval. We compare the sparseness-based method with the dispersal-based method, also proposed by Le Borgne et al., and the clustering-based method previously proposed by us. We show that the sparse filters and highly dispersed filters are quite different. In fact we show that highly dispersed filters tend to have lower sparseness. We also show that the sparse filters give better results compared to the highly dispersed filters when applied to texture retrieval. However the sparseness measures are calculated over filter response energies, making this method susceptible to choosing a redundant filter set. This issue is demonstrated and we show that ICF selected using our clustering-based method, which chooses a filter set with much lower redundancy, outperforms the sparse filters.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126909059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008082
Zahra M. Bagheri, S. Wiederman, B. Cazzolato, S. Grainger, D. O’Carroll
Many species of flying insects detect and chase prey or conspecifics within a visually cluttered surround, e.g. for predation, territorial or mating behavior. We modeled such detection and pursuit for small moving targets, and tested it within a closed-loop, virtual reality flight arena. Our model is inspired directly by electrophysiological recordings from 'small target motion detector' (STMD) neurons in the insect brain that are likely to underlie this behavioral task. The front-end uses a variant of a biologically inspired 'elementary' small target motion detector (ESTMD), elaborated to detect targets in natural scenes of both contrast polarities (i.e. both dark and light targets). We also include an additional model for the recently identified physiological 'facilitation' mechanism believed to form the basis for selective attention in insect STMDs, and quantify the improvement this provides for pursuit success and target discriminability over a range of target contrasts.
{"title":"A Biologically Inspired Facilitation Mechanism Enhances the Detection and Pursuit of Targets of Varying Contrast","authors":"Zahra M. Bagheri, S. Wiederman, B. Cazzolato, S. Grainger, D. O’Carroll","doi":"10.1109/DICTA.2014.7008082","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008082","url":null,"abstract":"Many species of flying insects detect and chase prey or conspecifics within a visually cluttered surround, e.g. for predation, territorial or mating behavior. We modeled such detection and pursuit for small moving targets, and tested it within a closed-loop, virtual reality flight arena. Our model is inspired directly by electrophysiological recordings from 'small target motion detector' (STMD) neurons in the insect brain that are likely to underlie this behavioral task. The front-end uses a variant of a biologically inspired 'elementary' small target motion detector (ESTMD), elaborated to detect targets in natural scenes of both contrast polarities (i.e. both dark and light targets). We also include an additional model for the recently identified physiological 'facilitation' mechanism believed to form the basis for selective attention in insect STMDs, and quantify the improvement this provides for pursuit success and target discriminability over a range of target contrasts.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127369766","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008112
Geoff Bull, Junbin Gao, M. Antolovich
Image segmentation seeks to partition the pixels in images into distinct regions to assist other image processing functions such as object recognition. Over the last few years dictionary learning methods have become very popular for image processing tasks such as denoising, and recently structured low rank dictionary learning has been shown to be capable of promising results for recognition tasks. This paper investigates the suitability of dictionary learning for image segmentation. A structured low rank dictionary learning algorithm is developed to segment images using compressed sensed features from image patches. To enable a supervised learning approach, classes of pixels in images are designated using training scribbles. A classifier is then learned from these training pixels and subsequently used to classify all other pixels in the images to form the segmentations. A number of dictionary learning models are compared together with K-means/nearest neighbour and support vector machine classifiers.
{"title":"Image Segmentation Using Dictionary Learning and Compressed Random Features","authors":"Geoff Bull, Junbin Gao, M. Antolovich","doi":"10.1109/DICTA.2014.7008112","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008112","url":null,"abstract":"Image segmentation seeks to partition the pixels in images into distinct regions to assist other image processing functions such as object recognition. Over the last few years dictionary learning methods have become very popular for image processing tasks such as denoising, and recently structured low rank dictionary learning has been shown to be capable of promising results for recognition tasks. This paper investigates the suitability of dictionary learning for image segmentation. A structured low rank dictionary learning algorithm is developed to segment images using compressed sensed features from image patches. To enable a supervised learning approach, classes of pixels in images are designated using training scribbles. A classifier is then learned from these training pixels and subsequently used to classify all other pixels in the images to form the segmentations. A number of dictionary learning models are compared together with K-means/nearest neighbour and support vector machine classifiers.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128032156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008102
B. Bozorgtabar, Roland Göcke
In this paper, we propose a discriminative multitask sparse learning scheme for object tracking in a particle filter framework. By representing each particle as a linear combination of adaptive dictionary templates, we utilise the correlations among different particles (tasks) to obtain a better representation and a more efficient scheme than learning each task individually. However, this model is completely generative and the designed tracker may not be robust enough to prevent the drifting problem in the presence of rapid appearance changes. In this paper, we use a Conditional Random Field (CRF) along with the multitask sparse model to extend our scheme to distinguish the object candidate from the background particle candidate. By this way, the number of particle samples is reduced significantly, while we make the tracker more robust. The proposed algorithm is evaluated on 11 challenging sequences and the results confirm the effectiveness of the approach and significantly outperforms the state-of-the-art trackers in terms of accuracy measures including the centre location error and the overlap ratio, respectively.
{"title":"Discriminative Multi-Task Sparse Learning for Robust Visual Tracking Using Conditional Random Field","authors":"B. Bozorgtabar, Roland Göcke","doi":"10.1109/DICTA.2014.7008102","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008102","url":null,"abstract":"In this paper, we propose a discriminative multitask sparse learning scheme for object tracking in a particle filter framework. By representing each particle as a linear combination of adaptive dictionary templates, we utilise the correlations among different particles (tasks) to obtain a better representation and a more efficient scheme than learning each task individually. However, this model is completely generative and the designed tracker may not be robust enough to prevent the drifting problem in the presence of rapid appearance changes. In this paper, we use a Conditional Random Field (CRF) along with the multitask sparse model to extend our scheme to distinguish the object candidate from the background particle candidate. By this way, the number of particle samples is reduced significantly, while we make the tracker more robust. The proposed algorithm is evaluated on 11 challenging sequences and the results confirm the effectiveness of the approach and significantly outperforms the state-of-the-art trackers in terms of accuracy measures including the centre location error and the overlap ratio, respectively.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125061315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008083
Md. Asikuzzaman, M. Alam, A. Lambert, M. Pickering
Piracy of a digital movie is a significant threat for movie studios and producers. Digital video watermarking is an important technique that can be used to protect the content. In existing watermarking algorithms, robustness to several attacks of the watermark has been improved. However, none of the existing techniques are robust to a combination of the common geometric distortions of scaling, rotation, and cropping with other attacks. In this paper, we propose a blind video watermarking algorithm where the watermark is embedded into both chrominance channels using a dual-tree complex wavelet transform. Embedding the watermark into the chrominance channels maintains the original video quality and the dual-tree complex wavelet transform ensures the robustness to geometric attacks due to its shift invariance characteristics. The watermark is extracted using the information from a single frame without using the original frame which makes this approach robust to temporal synchronization attacks such as frame dropping and frame rate change. This approach is also robust to downscaling in arbitrary resolution, aspect ratio change, compression, and camcording.
{"title":"A Blind and Robust Video Watermarking Scheme Using Chrominance Embedding","authors":"Md. Asikuzzaman, M. Alam, A. Lambert, M. Pickering","doi":"10.1109/DICTA.2014.7008083","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008083","url":null,"abstract":"Piracy of a digital movie is a significant threat for movie studios and producers. Digital video watermarking is an important technique that can be used to protect the content. In existing watermarking algorithms, robustness to several attacks of the watermark has been improved. However, none of the existing techniques are robust to a combination of the common geometric distortions of scaling, rotation, and cropping with other attacks. In this paper, we propose a blind video watermarking algorithm where the watermark is embedded into both chrominance channels using a dual-tree complex wavelet transform. Embedding the watermark into the chrominance channels maintains the original video quality and the dual-tree complex wavelet transform ensures the robustness to geometric attacks due to its shift invariance characteristics. The watermark is extracted using the information from a single frame without using the original frame which makes this approach robust to temporal synchronization attacks such as frame dropping and frame rate change. This approach is also robust to downscaling in arbitrary resolution, aspect ratio change, compression, and camcording.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114949247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008114
Fan Zhang, Yang Song, Sidong Liu, Sonia Pujol, R. Kikinis, D. Feng, Weidong (Tom) Cai
In this work, we propose a Latent Semantic Association Retrieval(LSAR) method to break the bottleneck of the low-level feature based medical image retrieval. The method constructs the high-level semantic correlations among patients based on the low-level feature set extracted from the images. Specifically, a Pair-LDA model is firstly designed to refine the topic generation process of traditional Latent Dirichlet Allocation (LDA), by generating the topics in a pair-wise context. Then, the latent association, called CCA-Correlation, is extracted to capture the correlations among the images in the Pair-LDA topic space based on Canonical Correlation Analysis (CCA). Finally, we calculate the similarity between images using the derived CCA-Correlation model and apply it to medical image retrieval. To evaluate the effectiveness of our method, we conduct the retrieval experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) baseline cohort with 331 subjects, and our method achieves good improvement compared to the state-of-the-art medical image retrieval methods. LSAR is independent on problem domain, thus can be generally applicable to other medical or general image analysis.
{"title":"Latent Semantic Association for Medical Image Retrieval","authors":"Fan Zhang, Yang Song, Sidong Liu, Sonia Pujol, R. Kikinis, D. Feng, Weidong (Tom) Cai","doi":"10.1109/DICTA.2014.7008114","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008114","url":null,"abstract":"In this work, we propose a Latent Semantic Association Retrieval(LSAR) method to break the bottleneck of the low-level feature based medical image retrieval. The method constructs the high-level semantic correlations among patients based on the low-level feature set extracted from the images. Specifically, a Pair-LDA model is firstly designed to refine the topic generation process of traditional Latent Dirichlet Allocation (LDA), by generating the topics in a pair-wise context. Then, the latent association, called CCA-Correlation, is extracted to capture the correlations among the images in the Pair-LDA topic space based on Canonical Correlation Analysis (CCA). Finally, we calculate the similarity between images using the derived CCA-Correlation model and apply it to medical image retrieval. To evaluate the effectiveness of our method, we conduct the retrieval experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) baseline cohort with 331 subjects, and our method achieves good improvement compared to the state-of-the-art medical image retrieval methods. LSAR is independent on problem domain, thus can be generally applicable to other medical or general image analysis.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"26 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120893411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008091
Jinglan Tian, Ling Li, Wanquan Liu
We work on the task of 2D articulated human pose tracking in monocular image sequences, an extremely challenging task due to background cluttering, variation in body appearance, occlusion and imaging conditions. Most of current approaches only deal with simple appearance and adjacent body part dependencies, especially the Gaussian tree-structured priors assumed over body part connections. Such prior makes the part connections independent to image evidence and in turn severely limits accuracy. Building on the successful pictorial structures model, we propose a novel framework combining an image-conditioned model that incorporates higher order dependencies of multiple body parts. In order to establish the conditioning variables, we employ the effective poselet features. In addition to this, we introduce a full body detector as the first step of our framework to reduce the search space for pose tracking. We evaluate our framework on two challenging image sequences and conduct a series of comparison experiments to compare the performance with another two approaches. The results illustrate that the proposed framework in this work outperforms the state-of-the-art 2D pose tracking systems.
{"title":"A Robust Framework for 2D Human Pose Tracking with Spatial and Temporal Constraints","authors":"Jinglan Tian, Ling Li, Wanquan Liu","doi":"10.1109/DICTA.2014.7008091","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008091","url":null,"abstract":"We work on the task of 2D articulated human pose tracking in monocular image sequences, an extremely challenging task due to background cluttering, variation in body appearance, occlusion and imaging conditions. Most of current approaches only deal with simple appearance and adjacent body part dependencies, especially the Gaussian tree-structured priors assumed over body part connections. Such prior makes the part connections independent to image evidence and in turn severely limits accuracy. Building on the successful pictorial structures model, we propose a novel framework combining an image-conditioned model that incorporates higher order dependencies of multiple body parts. In order to establish the conditioning variables, we employ the effective poselet features. In addition to this, we introduce a full body detector as the first step of our framework to reduce the search space for pose tracking. We evaluate our framework on two challenging image sequences and conduct a series of comparison experiments to compare the performance with another two approaches. The results illustrate that the proposed framework in this work outperforms the state-of-the-art 2D pose tracking systems.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"373 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131646295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008106
Jingsong Xu, Qiang Wu, Jian Zhang, B. Silk, Gia Thuan Ngo, Zhenmin Tang
People counting is a topic with various practical applications. Over the last decade, two general approaches have been proposed to tackle this problem: (a) counting based on individual human detection; (b)counting by measuring regression relation between the crowd density and number of people. Because the regression based method can avoid explicit people detection which faces several well-known challenges, it has been considered as a robust method particularly on a complicated environments. An efficient regression based method is proposed in this paper, which can be well adopted into any existing video surveillance system. It adopts color based segmentation to extract foreground regions in images. Regression is established based on the foreground density and the number of people. This method is fast and can deal with lighting condition changes. Experiments on public datasets and one captured dataset have shown the effectiveness and robustness of the method.
{"title":"Efficient People Counting with Limited Manual Interferences","authors":"Jingsong Xu, Qiang Wu, Jian Zhang, B. Silk, Gia Thuan Ngo, Zhenmin Tang","doi":"10.1109/DICTA.2014.7008106","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008106","url":null,"abstract":"People counting is a topic with various practical applications. Over the last decade, two general approaches have been proposed to tackle this problem: (a) counting based on individual human detection; (b)counting by measuring regression relation between the crowd density and number of people. Because the regression based method can avoid explicit people detection which faces several well-known challenges, it has been considered as a robust method particularly on a complicated environments. An efficient regression based method is proposed in this paper, which can be well adopted into any existing video surveillance system. It adopts color based segmentation to extract foreground regions in images. Regression is established based on the foreground density and the number of people. This method is fast and can deal with lighting condition changes. Experiments on public datasets and one captured dataset have shown the effectiveness and robustness of the method.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"76 7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127392613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-01DOI: 10.1109/DICTA.2014.7008108
Yan Zhao, Zhimin Gao, Lei Wang, Luping Zhou
Automatic identification of HEp-2 cell images has received an increasing research attention. Feature representations play a critical role in achieving good identification performance. Much recent work has focused on supervised feature learning. Typical methods consist of BoW model (based on hand-crafted features) and deep learning model (learning hierarchical features). However, these labels used in supervised feature learning are very labour-intensive and time-consuming. They are commonly manually annotated by specialists and very expensive to obtain. In this paper, we follow this fact and focus on unsupervised feature learning. We have verified and compared the features of these two typical models by clustering. Experimental results show the BoW model generally perform better than deep learning models. Also, we illustrate BoW model and deep learning models have complementarity properties.
{"title":"Experimental Study of Unsupervised Feature Learning for HEp-2 Cell Images Clustering","authors":"Yan Zhao, Zhimin Gao, Lei Wang, Luping Zhou","doi":"10.1109/DICTA.2014.7008108","DOIUrl":"https://doi.org/10.1109/DICTA.2014.7008108","url":null,"abstract":"Automatic identification of HEp-2 cell images has received an increasing research attention. Feature representations play a critical role in achieving good identification performance. Much recent work has focused on supervised feature learning. Typical methods consist of BoW model (based on hand-crafted features) and deep learning model (learning hierarchical features). However, these labels used in supervised feature learning are very labour-intensive and time-consuming. They are commonly manually annotated by specialists and very expensive to obtain. In this paper, we follow this fact and focus on unsupervised feature learning. We have verified and compared the features of these two typical models by clustering. Experimental results show the BoW model generally perform better than deep learning models. Also, we illustrate BoW model and deep learning models have complementarity properties.","PeriodicalId":146695,"journal":{"name":"2014 International Conference on Digital Image Computing: Techniques and Applications (DICTA)","volume":"151 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131746282","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：