Filters








3,177 Hits in 6.6 sec

Exploring Fisher vector and deep networks for action spotting

Zhe Wang, Limin Wang, Wenbin Du, Yu Qiao
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
However, our current deep network fails to yield better performance than our Fisher vector based approach and may need further exploration.  ...  For this reason, we submit the results obtained by our Fisher vector approach which achieves a Jaccard Index of 0.5385 and ranks the 1 st place in track 2.  ...  vector code based on the GMM [21] , (v) training SVMs [14] for classification.  ... 
doi:10.1109/cvprw.2015.7301330 dblp:conf/cvpr/WangWD015 fatcat:5c7akoaambajfppowkaz7viaaa

Deep Feature Representations for Variable-sized Regions of Interest in Breast Histopathology

Caner Mercan, Bulut Aygunes, Selim Aksoy, Ezgi Mercan, Linda G Shapiro, Donald L Weaver, Joann G Elmore
2020 IEEE journal of biomedical and health informatics  
Modeling variable-sized regions of interest (ROIs) in whole slide images using deep convolutional networks is a challenging task, as these networks typically require fixed-sized inputs that should contain  ...  sufficient structural and contextual information for classification.  ...  Whole slide imaging has aided these systems via digitization of glass slides into very high resolution images.  ... 
doi:10.1109/jbhi.2020.3036734 pmid:33166257 pmcid:PMC8274968 fatcat:xk7vu5jsd5eezhpr5onxnc24he

Image classification using object detectors

Thibaut Durand, Nicolas Thome, Matthieu Cord, Sandra Avila
2013 2013 IEEE International Conference on Image Processing  
Fisher Vectors and BossaNova. Our experiments carried out in the challenging PASCAL VOC 2007 dataset reveal outstanding performances.  ...  Image categorization is one of the most competitive topic in computer vision and image processing.  ...  However, we can highlight methods that enrich the BoW representation at the coding step, with sparse coding [3, 4] , or with a vectorial representation, as done in Fisher Vectors [5] .  ... 
doi:10.1109/icip.2013.6738894 dblp:conf/icip/DurandTCA13 fatcat:di7tsxzffncfjmzd7ukgusfh5y

Deep Region Hashing for Efficient Large-scale Instance Search from Images [article]

Jingkuan Song, Tao He, Lianli Gao, Xing Xu, Heng Tao Shen
2017 arXiv   pre-print
To tackle these issues, in this paper we propose an effective and efficient Deep Region Hashing (DRH) approach for large-scale INS using an image patch as the query.  ...  Specifically, DRH is an end-to-end deep neural network which consists of object proposal, feature extraction, and hash code generation.  ...  (a) is a query instance with a deep hash code representation Hq; (c) is a database image with a deep global region hash code Hg by taking the whole image as an region; (b) is an region pooled from (c)  ... 
arXiv:1701.07901v1 fatcat:e5ek7u2uajcc7gpvwnkii5q4cm

Image Retrieval Based on the Weighted and Regional Integration of CNN Features

2022 KSII Transactions on Internet and Information Systems  
Therefore, this paper proposes a feature weighting and region integration method for convolutional layer features to form global feature vectors and subsequently use them for image matching.  ...  Next, we integrate several regional eigenvectors that are processed by sliding windows into a global eigenvector.  ...  In the ILSVRC 2012 competition, Krizheysky et al. designed a deep convolutional network model called AlexNet [4] , which reduced the error rate of image classification from 26.2% to 15.3%, notably better  ... 
doi:10.3837/tiis.2022.03.008 fatcat:mwqhxgw4zjgktk52x6rfr34s4y

Efficient Maximum Appearance Search for Large-Scale Object Detection

Qiang Chen, Zheng Song, Rogerio Feris, Ankur Datta, Liangliang Cao, Zhongyang Huang, Shuicheng Yan
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
Our EMAS model consists of representing an image as an ensemble of densely sampled feature points with the proposed Pointwise Fisher Vector encoding method, so that the learnt discriminative scoring function  ...  5 seconds per image for the SUN09 dataset using a single CPU.  ...  We use the normalized Fisher Vector of the whole image (which can be easily produced from the PFVs) as features.  ... 
doi:10.1109/cvpr.2013.410 dblp:conf/cvpr/ChenSFDCHY13 fatcat:vq3qk6cui5hcjhara3l42ishqm

ChaLearn Looking at People Challenge 2014: Dataset and Results [chapter]

Sergio Escalera, Xavier Baró, Jordi Gonzàlez, Miguel A. Bautista, Meysam Madadi, Miguel Reyes, Víctor Ponce-López, Hugo J. Escalante, Jamie Shotton, Isabelle Guyon
2015 Lecture Notes in Computer Science  
For all the tracks, the goal was to perform user-independent recognition in sequences of continuous images using the overlapping Jaccard index as the evaluation measure.  ...  Results achieved an overlapping accuracy about 0.20 and 0.50 for pose recovery and action/interaction spotting, showing still much margin for improvement, meanwhile an overlapping about 0.85 was achieved  ...  Special thanks to Pau Rodríguez for annotating part of the multi-modal gestures. We thank Microsoft Codalab submission website and researchers who joined the PC and reviewed for the workshop.  ... 
doi:10.1007/978-3-319-16178-5_32 fatcat:arlegvujt5c2fb3fmou4fx5iwu

A Comprehensive Solution for Deep-Learning Based Cargo Inspection to Discriminate Goods in Containers

Jiahang Che, Yuxiang Xing, Li Zhang
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
By employing deep learning, we train a triplet network for finegrained image categorization.  ...  In this work, we attempt to classify commodities in containers with HS(harmonized system) codes, which is a challenging task due to the large number of categories in HS codes and its hierarchical structure  ...  We would like to thank Jian Zhang for his insightful suggestions to this work. Also we are thankful to our colleagues and especially grateful to Qiang Li and Gang Fu for their discussions.  ... 
doi:10.1109/cvprw.2018.00166 dblp:conf/cvpr/CheXZ18 fatcat:d3tom4ibtbhwpn2smnqtytwhhm

Convolution Neural Network Based Deep Feature Fusion for Palmprint and Handvein

2019 International Journal of Engineering and Advanced Technology  
A distinctive Deep Convolution Neural Network (CNN) outline architecture model is presented in this work which proficiently signify complex image features.  ...  feature extraction and classification outcomes.  ...  As a final point, softmax layer is used for fruitful classification of the input sliding window image. IV.  ... 
doi:10.35940/ijeat.f9217.109119 fatcat:xvlktgenwrgz3jbx4zyy4fszui

Recognizing Thousands of Legal Entities through Instance-based Visual Classification

Valentin Leveau, Alexis Joly, Olivier Buisson, Pierre Letessier, Patrick Valduriez
2014 Proceedings of the ACM International Conference on Multimedia - MM '14  
This paper considers the problem of recognizing legal entities in visual contents in a similar way to named-entity recognizers for text documents.  ...  We introduce a new geometrically-consistent instance-based classification method that is shown to outperform state-of-the-art techniques on several challenging datasets while being much more scalable.  ...  They report some consistent performance improvements over several state-of-the-art classification methods (including Fisher Vectors [14] ).  ... 
doi:10.1145/2647868.2655038 dblp:conf/mm/LeveauJBLV14 fatcat:nocedugyrfh2tg4yzhtcijd2vi

SRI-Sarnoff AURORA System at TRECVID 2014 Multimedia Event Detection and Recounting

Hui Cheng, Jingen Liu, Ishani Chakraborty, Guang Chen, Qiguang Liu, Mohamed Elhoseiny, Gary Gan, Ajay Divakaran, Harpreet S. Sawhney, James Allan, John Foley, Mubarak Shah (+3 others)
2014 TREC Video Retrieval Evaluation  
Aurora system extracts multi-modality features including motion features, static image feature, and audio features from videos, and represents a video with Bag-of-Word (BOW) and Fisher Vector model.  ...  The deep-learning features achieve good performance for MED, but they are not the right features for MER.  ...  Government is authorized to reproduce and distribute reprints for Governmental purposes not with-standing any copyright annotation thereon.  ... 
dblp:conf/trecvid/ChengLCCLEGDSAF14 fatcat:gag3qmwubfa2difyos4g3dylim

Deep Patch Learning for Weakly Supervised Object Classification and Discovery [article]

Peng Tang, Xinggang Wang, Zilong Huang, Xiang Bai, Wenyu Liu
2017 arXiv   pre-print
Patch-level image representation is very important for object classification and detection, since it is robust to spatial transformation, scale variation, and cluttered background.  ...  The network processes the two tasks object classification and discovery jointly, and shares hierarchical deep features.  ...  (a) and (b) produce a score vector per-image for classification and only require image-level annotations for training.  ... 
arXiv:1705.02429v1 fatcat:wtdiqqifkngzdcajpcxupkodbq

A Novel SAR Image Target Recognition Algorithm under Big Data Analysis

Xiang Chen, Xing Wang, You Chen, Haihan Wang, Thippa Reddy G
2021 Wireless Communications and Mobile Computing  
Finally, the network model modified by softmax cross-entropy loss and Fisher loss is used for automatic target recognition.  ...  Based on the MSTAR data set, two scene graphs containing the target synthesized by the background image and the target slice are used for experiments.  ...  For the classification and recognition task of natural images, researchers have proposed some very typical deep learning network models, such as AlexNet [6] , VGG [7] , Googlenet [8] , ResNet [9] ,  ... 
doi:10.1155/2021/4556157 fatcat:dw3cppfdurhexlf73ckqoo63om

Deep Learning for Scene Classification: A Survey [article]

Delu Zeng, Minyu Liao, Mohammad Tavakolian, Yulan Guo, Bolei Zhou, Dewen Hu, Matti Pietikäinen, Li Liu
2021 arXiv   pre-print
Scene classification, aiming at classifying a scene image to one of the predefined scene categories by comprehending the entire image, is a longstanding, fundamental and challenging problem in computer  ...  To help researchers master needed advances in this field, the goal of this paper is to provide a comprehensive survey of recent achievements in scene classification using deep learning.  ...  ACKNOWLEDGMENTS The authors would like to thank the pioneer researchers in scene classification and other related fields. This work was supported in part by grants from National Science  ... 
arXiv:2101.10531v2 fatcat:hwqw5so46ngxdlnfw7zynmpu6m

ChaLearn Looking at People 2015 challenges: Action spotting and cultural event recognition

Xavier Baro, Jordi Gonzalez, Junior Fabian, Miguel A. Bautista, Marc Oliu, Hugo Jair Escalante, Isabelle Guyon, Sergio Escalera
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Tesla K40 GPU used for creating the baseline of the Cultural Event Recognition track.  ...  For each sliding window, the authors used the pooled Fisher Vector as representation and fed it into the SVM classifier for action recognition. A summary of this method is shown in Figure 5 .  ...  Then, for each kind of descriptor, the participants trained a GMM and used Fisher vector to transform these descriptors into a high dimensional super vector space.  ... 
doi:10.1109/cvprw.2015.7301329 dblp:conf/cvpr/BaroGFBOEGE15 fatcat:factqdhbrfd5fgzuepo4aovwoa
« Previous Showing results 1 — 15 out of 3,177 results