Filters








46,184 Hits in 5.3 sec

Fully-adaptive Feature Sharing in Multi-Task Networks with Applications in Person Attribute Classification [article]

Yongxi Lu, Abhishek Kumar, Shuangfei Zhai, Yu Cheng, Tara Javidi, Rogerio Feris
2016 arXiv   pre-print
In the context of deep neural networks, this idea is often realized by hand-designed network architectures with layers that are shared across tasks and branches that encode task-specific features.  ...  Our Extensive evaluation on person attributes classification tasks involving facial and clothing attributes suggests that the models produced by the proposed method are fast, compact and can closely match  ...  Multi-task networks have been used with base layers that are shared across all attributes, and branches to encode task-specific features for each attribute category [13, 37] .  ... 
arXiv:1611.05377v1 fatcat:ocoaqzzfynfcxbhmtrmslrexra

Fully-Adaptive Feature Sharing in Multi-Task Networks with Applications in Person Attribute Classification

Yongxi Lu, Abhishek Kumar, Shuangfei Zhai, Yu Cheng, Tara Javidi, Rogerio Feris
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In the context of deep neural networks, this idea is often realized by handdesigned network architectures with layers that are shared across tasks and branches that encode task-specific features.  ...  Our approach starts with a thin multi-layer network and dynamically widens it in a greedy manner during training.  ...  It thus can be applied to multi-task problem beyond person attribute classifications in future works.  ... 
doi:10.1109/cvpr.2017.126 dblp:conf/cvpr/LuKZCJF17 fatcat:j7j2ieuyijd6vhuil374rtz37a

Two-Stream Multi-Task Network for Fashion Recognition [article]

Peizhao Li, Yanjing Li, Xiaolong Jiang, Xiantong Zhen
2019 arXiv   pre-print
In this paper, we present a two-stream multi-task network for fashion recognition.  ...  To handle these challenges, we formulate fashion recognition into a multi-task learning problem, including landmark detection, category and attribute classifications, and solve it with the proposed deep  ...  We replace the last two fully connected layers in fashion classification network with a two-branch fully connected layer, in which way accomplish the parameter sharing for MTL.  ... 
arXiv:1901.10172v3 fatcat:iczeovc2i5gd7dytgciawrw36a

Multi-task Mid-level Feature Alignment Network for Unsupervised Cross-Dataset Person Re-Identification [article]

Shan Lin, Haoliang Li, Chang-Tsun Li, Alex Chichung Kot
2018 arXiv   pre-print
To overcome this limitation, we develop a novel unsupervised Multi-task Mid-level Feature Alignment (MMFA) network for the unsupervised cross-dataset person re-identification task.  ...  the attribute learning task with a cross-dataset mid-level feature alignment regularisation term.  ...  Multi-task Supervised Classification for Feature Learning The view-invariant feature representations are learned from a multi-task identity and attribute classification training.  ... 
arXiv:1807.01440v2 fatcat:26gw74g7nfblvf53edk32dc4ju

Adaptively Weighted Multi-task Deep Network for Person Attribute Classification

Keke He, Zhanxiong Wang, Yanwei Fu, Rui Feng, Yu-Gang Jiang, Xiangyang Xue
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
To solve these problems, we present a weighted multi-task deep convolutional neural network for person attribute analysis.  ...  KEYWORDS facial attribute analysis, person attribute analysis, deep learning, multi-task learning ACM Reference Format:  ...  CONCLUSIONS In this paper, we propose a novel Adaptively Weighted Multi-task Deep Convolutional Neural Network to learn person attributes.  ... 
doi:10.1145/3123266.3123424 dblp:conf/mm/HeWFFJX17 fatcat:adgmeqs6gvexvnr77gtuid3imm

Pedestrian Attribute Recognition: A Survey [article]

Xiao Wang, Shaofei Zheng, Rui Yang, Bin Luo, Jin Tang
2019 arXiv   pre-print
Recognizing pedestrian attributes is an important task in computer vision community due to it plays an important role in video surveillance. Many algorithms has been proposed to handle this task.  ...  Thirdly, we analyse the concept of multi-task learning and multi-label learning, and also explain the relations between these two learning algorithms and pedestrian attribute recognition.  ...  It is worthy to note that all attributes in the same group share the same fully connected feature.  ... 
arXiv:1901.07474v1 fatcat:h5krexsotbecvlwi2w4uw2y3ay

Deep Multi-task Multi-label CNN for Effective Facial Attribute Classification [article]

Longbiao Mao, Yan Yan, Jing-Hao Xue, Hanzi Wang
2020 arXiv   pre-print
Facial Attribute Classification (FAC) has attracted increasing attention in computer vision and pattern recognition.  ...  The inherent dependencies between these tasks are not fully exploited.  ...  [30] develop a multi-task attention network, which automatically learns both task-shared and task-specific features in an end-to-end manner, for MTL.  ... 
arXiv:2002.03683v1 fatcat:wvrup4lxnje7nki47fikzmf5di

Survey on Reliable Deep Learning-Based Person Re-Identification Models: Are We There Yet? [article]

Bahram Lavi, Ihsan Ullah, Mehdi Fatan, Anderson Rocha
2020 arXiv   pre-print
Person re-identification (PReID) is one of the most critical problems in IVS, and it consists of recognizing whether or not an individual has already been observed over a camera in a network.  ...  Given the importance and wide range of applications of re-identification solutions, our objective herein is to discuss the work carried out in the area and come up with a survey of state-of-the-art DNN  ...  and Interpretation of Events" with grant number 18/05668-3.  ... 
arXiv:2005.00355v1 fatcat:5msfk3apirg6vnpja52iw2qd3e

Unsupervised Person Re-Identification: A Systematic Survey of Challenges and Solutions [article]

Xiangtan Lin and Pengzhen Ren and Chung-Hsing Yeh and Lina Yao and Andy Song and Xiaojun Chang
2021 arXiv   pre-print
person feature learning; 2) learning discriminative person features with pseudo-supervision; 3) learning cross-camera invariant person feature, and 4) the domain shift between datasets.  ...  Existing person Re-ID surveys have focused on supervised methods from classifications and applications but lack detailed discussion on how the person Re-ID solutions address the underlying challenges.  ...  JVTC [91] tackles the domain gap challenge by enforcing visual and temporal consistency in a multi-class classification task.  ... 
arXiv:2109.06057v2 fatcat:epfow7w3trevff5iku2uvb4ov4

Cross-stitch Networks for Multi-task Learning [article]

Ishan Misra and Abhinav Shrivastava and Abhinav Gupta and Martial Hebert
2016 arXiv   pre-print
Multi-task learning in Convolutional Networks has displayed remarkable success in the field of recognition.  ...  A network with cross-stitch units can learn an optimal combination of shared and task-specific representations.  ...  This work was supported in part by ONR MURI N000141612007 and the US Army Research Laboratory (ARL) under the CTA program (Agreement W911NF-10-2-0016). AS was supported by the MSR fellowship.  ... 
arXiv:1604.03539v1 fatcat:qjhwexuju5fhjg4anqoj7tmmgu

Deep domain adaptation for describing people based on fine-grained clothing attributes

Qiang Chen, Junshi Huang, Rogerio Feris, Lisa M Brown, Jian Dong, Shuicheng Yan
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In order to bridge this gap, we propose a novel double-path deep domain adaptation network to model the data from the two domains jointly.  ...  Several alignment cost layers placed inbetween the two columns ensure the consistency of the two domain features and the feasibility to predict unseen attribute categories in one of the domains.  ...  Introduction Describing people in detail is an important task for many applications.  ... 
doi:10.1109/cvpr.2015.7299169 dblp:conf/cvpr/ChenHFBDY15 fatcat:uqxdx75jt5bldd7kruaoovbnoi

A Survey of Deep Facial Attribute Analysis [article]

Xin Zheng, Yanqing Guo, Huaibo Huang, Yi Li, Ran He
2019 arXiv   pre-print
Furthermore, several additional facial attribute related issues are introduced, as well as relevant real-world applications.  ...  Deep learning based facial attribute analysis consists of two basic sub-issues: facial attribute estimation (FAE), which recognizes whether facial attributes are present in given images, and facial attribute  ...  The proposed network learns shared features in a fully adaptive way, where the core idea is incrementally widening the current design in a layer-wise manner.  ... 
arXiv:1812.10265v3 fatcat:tezgo2angvfefbttuoodnss6t4

Deep fusion of visual signatures for client-server facial analysis

Binod Bhattarai, Gaurav Sharma, Frederic Jurie
2016 Proceedings of the Tenth Indian Conference on Computer Vision, Graphics and Image Processing - ICVGIP '16  
the face e.g. is the person male or female, is she/he bald, does he have a mustache, etc.  ...  The challenge addressed in this paper is to design a common universal representation such that a single merged signature is transmitted to the server, whatever be the type and number of features computed  ...  Acknowledgments This project is funded in part by the ANR (grant ANR-12-SECU-0005).  ... 
doi:10.1145/3009977.3010062 dblp:conf/icvgip/Bhattarai0J16 fatcat:nxeljboim5c2jghjmpn6sbc5fu

Deep fusion of visual signatures for client-server facial analysis [article]

Binod Bhattarai, Gaurav Sharma, Frederic Jurie
2016 arXiv   pre-print
the face e.g. is the person male or female, is she/he bald, does he have a mustache, etc.  ...  The challenge addressed in this paper is to design a common universal representation such that a single merged signature is transmitted to the server, whatever be the type and number of features computed  ...  Acknowledgments This project is funded in part by the ANR (grant ANR-12-SECU-0005).  ... 
arXiv:1611.00142v2 fatcat:ltpotdz7fjherahjkoaya26wie

Augmenting Image Aesthetic Assessment with Diverse Deep Features

Rui Lin
2021 2021 4th Artificial Intelligence and Cloud Computing Conference  
With the increasing prevalence of digital images, automatically assessing the aesthetic quality of photos could benefit many realworld applications.  ...  While many previous methods have produced binary classification results, this paper proposes a model to produce regression results with high accuracy.  ...  It was replaced by a layer with 2 outputs for the classification task and a layer with one output for the regression task.  ... 
doi:10.1145/3508259.3508264 fatcat:ibsbli4xqvc6rjtkhcqi4fletq
« Previous Showing results 1 — 15 out of 46,184 results