Pub Date : 2020-06-01DOI: 10.1109/CVPR42600.2020.00353
Miao Zhang, Weisong Ren, Yongri Piao, Zhengkun Rong, Huchuan Lu
Depth data containing a preponderance of discriminative power in location have been proven beneficial for accurate saliency prediction. However, RGB-D saliency detection methods are also negatively influenced by randomly distributed erroneous or missing regions on the depth map or along the object boundaries. This offers the possibility of achieving more effective inference by well designed models. In this paper, we propose a new framework for accurate RGB-D saliency detection taking account of local and global complementarities from two modalities. This is achieved by designing a complimentary interaction model discriminative enough to simultaneously select useful representation from RGB and depth data, and meanwhile to refine the object boundaries. Moreover, we proposed a compensation-aware loss to further process the information not being considered in the complimentary interaction model, leading to improvement of the generalization ability for challenging scenes. Experiments on six public datasets show that our method outperforms18state-of-the-art methods.
{"title":"Select, Supplement and Focus for RGB-D Saliency Detection","authors":"Miao Zhang, Weisong Ren, Yongri Piao, Zhengkun Rong, Huchuan Lu","doi":"10.1109/CVPR42600.2020.00353","DOIUrl":"https://doi.org/10.1109/CVPR42600.2020.00353","url":null,"abstract":"Depth data containing a preponderance of discriminative power in location have been proven beneficial for accurate saliency prediction. However, RGB-D saliency detection methods are also negatively influenced by randomly distributed erroneous or missing regions on the depth map or along the object boundaries. This offers the possibility of achieving more effective inference by well designed models. In this paper, we propose a new framework for accurate RGB-D saliency detection taking account of local and global complementarities from two modalities. This is achieved by designing a complimentary interaction model discriminative enough to simultaneously select useful representation from RGB and depth data, and meanwhile to refine the object boundaries. Moreover, we proposed a compensation-aware loss to further process the information not being considered in the complimentary interaction model, leading to improvement of the generalization ability for challenging scenes. Experiments on six public datasets show that our method outperforms18state-of-the-art methods.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"27 1","pages":"3469-3478"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83986222","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 : 2020-06-01DOI: 10.1109/CVPR42600.2020.00267
Yining Lang, Yuan He, Fan Yang, Jianfeng Dong, Hui Xue
With the rapid growth of e-commerce and the popularity of online shopping, fashion retrieval has received considerable attention in the computer vision community. Different from the existing works that mainly focus on identical or similar fashion item retrieval, in this paper, we aim to study the plagiarized clothes retrieval which is somewhat ignored in the academic community while itself has great application value. One of the key challenges is that plagiarized clothes are usually modified in a certain region on the original design to escape the supervision by traditional retrieval methods. To relieve it, we propose a novel network named Plagiarized-Search-Net (PS-Net) based on regional representation, where we utilize the landmarks to guide the learning of regional representations and compare fashion items region by region. Besides, we propose a new dataset named Plagiarized Fashion for plagiarized clothes retrieval, which provides a meaningful complement to the existing fashion retrieval field. Experiments on Plagiarized Fashion dataset verify that our approach is superior to other instance-level counterparts for plagiarized clothes retrieval, showing a promising result for original design protection. Moreover, our PS-Net can also be adapted to traditional fashion retrieval and landmark estimation tasks and achieves the state-of-the-art performance on the DeepFashion and DeepFashion2 datasets.
{"title":"Which Is Plagiarism: Fashion Image Retrieval Based on Regional Representation for Design Protection","authors":"Yining Lang, Yuan He, Fan Yang, Jianfeng Dong, Hui Xue","doi":"10.1109/CVPR42600.2020.00267","DOIUrl":"https://doi.org/10.1109/CVPR42600.2020.00267","url":null,"abstract":"With the rapid growth of e-commerce and the popularity of online shopping, fashion retrieval has received considerable attention in the computer vision community. Different from the existing works that mainly focus on identical or similar fashion item retrieval, in this paper, we aim to study the plagiarized clothes retrieval which is somewhat ignored in the academic community while itself has great application value. One of the key challenges is that plagiarized clothes are usually modified in a certain region on the original design to escape the supervision by traditional retrieval methods. To relieve it, we propose a novel network named Plagiarized-Search-Net (PS-Net) based on regional representation, where we utilize the landmarks to guide the learning of regional representations and compare fashion items region by region. Besides, we propose a new dataset named Plagiarized Fashion for plagiarized clothes retrieval, which provides a meaningful complement to the existing fashion retrieval field. Experiments on Plagiarized Fashion dataset verify that our approach is superior to other instance-level counterparts for plagiarized clothes retrieval, showing a promising result for original design protection. Moreover, our PS-Net can also be adapted to traditional fashion retrieval and landmark estimation tasks and achieves the state-of-the-art performance on the DeepFashion and DeepFashion2 datasets.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"91 1","pages":"2592-2601"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80529364","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}
Current face detectors utilize anchors to frame a multi-task learning problem which combines classification and bounding box regression. Effective anchor design and anchor matching strategy enable face detectors to localize faces under large pose and scale variations. However, we observe that, more than 80% correctly predicted bounding boxes are regressed from the unmatched anchors (the IoUs between anchors and target faces are lower than a threshold) in the inference phase. It indicates that these unmatched anchors perform excellent regression ability, but the existing methods neglect to learn from them. In this paper, we propose an Online High-quality Anchor Mining Strategy (HAMBox), which explicitly helps outer faces compensate with high-quality anchors. Our proposed HAMBox method could be a general strategy for anchor-based single-stage face detection. Experiments on various datasets, including WIDER FACE, FDDB, AFW and PASCAL Face, demonstrate the superiority of the proposed method.
{"title":"HAMBox: Delving Into Mining High-Quality Anchors on Face Detection","authors":"Yang Liu, Xu Tang, Junyu Han, Jingtuo Liu, Dinger Rui, Xiang Wu","doi":"10.1109/cvpr42600.2020.01306","DOIUrl":"https://doi.org/10.1109/cvpr42600.2020.01306","url":null,"abstract":"Current face detectors utilize anchors to frame a multi-task learning problem which combines classification and bounding box regression. Effective anchor design and anchor matching strategy enable face detectors to localize faces under large pose and scale variations. However, we observe that, more than 80% correctly predicted bounding boxes are regressed from the unmatched anchors (the IoUs between anchors and target faces are lower than a threshold) in the inference phase. It indicates that these unmatched anchors perform excellent regression ability, but the existing methods neglect to learn from them. In this paper, we propose an Online High-quality Anchor Mining Strategy (HAMBox), which explicitly helps outer faces compensate with high-quality anchors. Our proposed HAMBox method could be a general strategy for anchor-based single-stage face detection. Experiments on various datasets, including WIDER FACE, FDDB, AFW and PASCAL Face, demonstrate the superiority of the proposed method.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"69 1","pages":"13043-13051"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80545054","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 : 2020-06-01DOI: 10.1109/CVPR42600.2020.00868
Weibin Wu, Yuxin Su, Xixian Chen, Shenglin Zhao, Irwin King, M. Lyu, Yu-Wing Tai
With the growing prevalence of convolutional neural networks (CNNs), there is an urgent demand to explain their behaviors. Global explanations contribute to understanding model predictions on a whole category of samples, and thus have attracted increasing interest recently. However, existing methods overwhelmingly conduct separate input attribution or rely on local approximations of models, making them fail to offer faithful global explanations of CNNs. To overcome such drawbacks, we propose a novel two-stage framework, Attacking for Interpretability (AfI), which explains model decisions in terms of the importance of user-defined concepts. AfI first conducts a feature occlusion analysis, which resembles a process of attacking models to derive the category-wide importance of different features. We then map the feature importance to concept importance through ad-hoc semantic tasks. Experimental results confirm the effectiveness of AfI and its superiority in providing more accurate estimations of concept importance than existing proposals.
{"title":"Towards Global Explanations of Convolutional Neural Networks With Concept Attribution","authors":"Weibin Wu, Yuxin Su, Xixian Chen, Shenglin Zhao, Irwin King, M. Lyu, Yu-Wing Tai","doi":"10.1109/CVPR42600.2020.00868","DOIUrl":"https://doi.org/10.1109/CVPR42600.2020.00868","url":null,"abstract":"With the growing prevalence of convolutional neural networks (CNNs), there is an urgent demand to explain their behaviors. Global explanations contribute to understanding model predictions on a whole category of samples, and thus have attracted increasing interest recently. However, existing methods overwhelmingly conduct separate input attribution or rely on local approximations of models, making them fail to offer faithful global explanations of CNNs. To overcome such drawbacks, we propose a novel two-stage framework, Attacking for Interpretability (AfI), which explains model decisions in terms of the importance of user-defined concepts. AfI first conducts a feature occlusion analysis, which resembles a process of attacking models to derive the category-wide importance of different features. We then map the feature importance to concept importance through ad-hoc semantic tasks. Experimental results confirm the effectiveness of AfI and its superiority in providing more accurate estimations of concept importance than existing proposals.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"86 1","pages":"8649-8658"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80790460","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 : 2020-06-01DOI: 10.1109/cvpr42600.2020.00726
André Mateus, S. Ramalingam, Pedro Miraldo
We explore the possibility of using line intersection constraints for 3D scan registration. Typical 3D registration algorithms exploit point and plane correspondences, while line intersection constraints have not been used in the context of 3D scan registration before. Constraints from a match of pairs of intersecting lines in two 3D scans can be seen as two 3D line intersections, a plane correspondence, and a point correspondence. In this paper, we present minimal solvers that combine these different type of constraints: 1) three line intersections and one point match; 2) one line intersection and two point matches; 3) three line intersections and one plane match; 4) one line intersection and two plane matches; and 5) one line intersection, one point match, and one plane match. To use all the available solvers, we present a hybrid RANSAC loop. We propose a non-linear refinement technique using all the inliers obtained from the RANSAC. Vast experiments with simulated data and two real-data data-sets show that the use of these features and the combined solvers improve the accuracy. The code is available.
{"title":"Minimal Solvers for 3D Scan Alignment With Pairs of Intersecting Lines","authors":"André Mateus, S. Ramalingam, Pedro Miraldo","doi":"10.1109/cvpr42600.2020.00726","DOIUrl":"https://doi.org/10.1109/cvpr42600.2020.00726","url":null,"abstract":"We explore the possibility of using line intersection constraints for 3D scan registration. Typical 3D registration algorithms exploit point and plane correspondences, while line intersection constraints have not been used in the context of 3D scan registration before. Constraints from a match of pairs of intersecting lines in two 3D scans can be seen as two 3D line intersections, a plane correspondence, and a point correspondence. In this paper, we present minimal solvers that combine these different type of constraints: 1) three line intersections and one point match; 2) one line intersection and two point matches; 3) three line intersections and one plane match; 4) one line intersection and two plane matches; and 5) one line intersection, one point match, and one plane match. To use all the available solvers, we present a hybrid RANSAC loop. We propose a non-linear refinement technique using all the inliers obtained from the RANSAC. Vast experiments with simulated data and two real-data data-sets show that the use of these features and the combined solvers improve the accuracy. The code is available.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"90 1","pages":"7232-7242"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80321975","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 : 2020-06-01DOI: 10.1109/CVPR42600.2020.01187
Yuqing Zhu, Xiang Yu, Manmohan Chandraker, Yu-Xiang Wang
With increasing ethical and legal concerns on privacy for deep models in visual recognition, differential privacy has emerged as a mechanism to disguise membership of sensitive data in training datasets. Recent methods like Private Aggregation of Teacher Ensembles (PATE) leverage a large ensemble of teacher models trained on disjoint subsets of private data, to transfer knowledge to a student model with privacy guarantees. However, labeled vision data is often expensive and datasets, when split into many disjoint training sets, lead to signiï¬cantly sub-optimal accuracy and thus hardly sustain good privacy bounds. We propose a practically data-efï¬cient scheme based on private release of k-nearest neighbor (kNN) queries, which altogether avoids splitting the training dataset. Our approach allows the use of privacy-ampliï¬cation by subsampling and iterative reï¬nement of the kNN feature embedding. We rigorously analyze the theoretical properties of our method and demonstrate strong experimental performance on practical computer vision datasets for face attribute recognition and person reidentiï¬cation. In particular, we achieve comparable or better accuracy than PATE while reducing more than 90% of the privacy loss, thereby providing the “most practical method to-date” for private deep learning in computer vision.
{"title":"Private-kNN: Practical Differential Privacy for Computer Vision","authors":"Yuqing Zhu, Xiang Yu, Manmohan Chandraker, Yu-Xiang Wang","doi":"10.1109/CVPR42600.2020.01187","DOIUrl":"https://doi.org/10.1109/CVPR42600.2020.01187","url":null,"abstract":"With increasing ethical and legal concerns on privacy for deep models in visual recognition, differential privacy has emerged as a mechanism to disguise membership of sensitive data in training datasets. Recent methods like Private Aggregation of Teacher Ensembles (PATE) leverage a large ensemble of teacher models trained on disjoint subsets of private data, to transfer knowledge to a student model with privacy guarantees. However, labeled vision data is often expensive and datasets, when split into many disjoint training sets, lead to signiï¬cantly sub-optimal accuracy and thus hardly sustain good privacy bounds. We propose a practically data-efï¬cient scheme based on private release of k-nearest neighbor (kNN) queries, which altogether avoids splitting the training dataset. Our approach allows the use of privacy-ampliï¬cation by subsampling and iterative reï¬nement of the kNN feature embedding. We rigorously analyze the theoretical properties of our method and demonstrate strong experimental performance on practical computer vision datasets for face attribute recognition and person reidentiï¬cation. In particular, we achieve comparable or better accuracy than PATE while reducing more than 90% of the privacy loss, thereby providing the “most practical method to-date” for private deep learning in computer vision.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"88 1","pages":"11851-11859"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75866451","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 : 2020-06-01DOI: 10.1109/cvpr42600.2020.00936
Vignesh Ramanathan, Rui Wang, D. Mahajan
Large detection datasets have a long tail of lowshot classes with very few bounding box annotations. We wish to improve detection for lowshot classes with weakly labelled web-scale datasets only having image-level labels. This requires a detection framework that can be jointly trained with limited number of bounding box annotated images and large number of weakly labelled images. Towards this end, we propose a modification to the FRCNN model to automatically infer label assignment for objects proposals from weakly labelled images during training. We pose this label assignment as a Linear Program with constraints on the number and overlap of object instances in an image. We show that this can be solved efficiently during training for weakly labelled images. Compared to just training with few annotated examples, augmenting with weakly labelled examples in our framework provides significant gains. We demonstrate this on the LVIS dataset 3.5 gain in AP as well as different lowshot variants of the COCO dataset. We provide a thorough analysis of the effect of amount of weakly labelled and fully labelled data required to train the detection model. Our DLWL framework can also outperform self-supervised baselines like omni-supervision for lowshot classes.
{"title":"DLWL: Improving Detection for Lowshot Classes With Weakly Labelled Data","authors":"Vignesh Ramanathan, Rui Wang, D. Mahajan","doi":"10.1109/cvpr42600.2020.00936","DOIUrl":"https://doi.org/10.1109/cvpr42600.2020.00936","url":null,"abstract":"Large detection datasets have a long tail of lowshot classes with very few bounding box annotations. We wish to improve detection for lowshot classes with weakly labelled web-scale datasets only having image-level labels. This requires a detection framework that can be jointly trained with limited number of bounding box annotated images and large number of weakly labelled images. Towards this end, we propose a modification to the FRCNN model to automatically infer label assignment for objects proposals from weakly labelled images during training. We pose this label assignment as a Linear Program with constraints on the number and overlap of object instances in an image. We show that this can be solved efficiently during training for weakly labelled images. Compared to just training with few annotated examples, augmenting with weakly labelled examples in our framework provides significant gains. We demonstrate this on the LVIS dataset 3.5 gain in AP as well as different lowshot variants of the COCO dataset. We provide a thorough analysis of the effect of amount of weakly labelled and fully labelled data required to train the detection model. Our DLWL framework can also outperform self-supervised baselines like omni-supervision for lowshot classes.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"4 1","pages":"9339-9349"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81371611","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 : 2020-06-01DOI: 10.1109/cvpr42600.2020.00912
Xiang Gu, Jian Sun, Zongben Xu
Adversarial domain adaptation (DA) has been an effective approach for learning domain-invariant features by adversarial training. In this paper, we propose a novel adversarial DA approach completely defined in spherical feature space, in which we define spherical classifier for label prediction and spherical domain discriminator for discriminating domain labels. To utilize pseudo-label robustly, we develop a robust pseudo-label loss in the spherical feature space, which weights the importance of estimated labels of target data by posterior probability of correct labeling, modeled by Gaussian-uniform mixture model in spherical feature space. Extensive experiments show that our method achieves state-of-the-art results, and also confirm effectiveness of spherical classifier, spherical discriminator and spherical robust pseudo-label loss.
{"title":"Spherical Space Domain Adaptation With Robust Pseudo-Label Loss","authors":"Xiang Gu, Jian Sun, Zongben Xu","doi":"10.1109/cvpr42600.2020.00912","DOIUrl":"https://doi.org/10.1109/cvpr42600.2020.00912","url":null,"abstract":"Adversarial domain adaptation (DA) has been an effective approach for learning domain-invariant features by adversarial training. In this paper, we propose a novel adversarial DA approach completely defined in spherical feature space, in which we define spherical classifier for label prediction and spherical domain discriminator for discriminating domain labels. To utilize pseudo-label robustly, we develop a robust pseudo-label loss in the spherical feature space, which weights the importance of estimated labels of target data by posterior probability of correct labeling, modeled by Gaussian-uniform mixture model in spherical feature space. Extensive experiments show that our method achieves state-of-the-art results, and also confirm effectiveness of spherical classifier, spherical discriminator and spherical robust pseudo-label loss.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"16 1","pages":"9098-9107"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80893537","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 : 2020-06-01DOI: 10.1109/CVPR42600.2020.00266
Wenkai Dong, Zhaoxiang Zhang, Chunfeng Song, T. Tan
Person detection networks have been widely used in person search. These detectors discriminate persons from the background and generate proposals of all the persons from a gallery of scene images for each query. However, such a large number of proposals have a negative influence on the following identity matching process because many distractors are involved. In this paper, we propose a new detection network for person search, named Instance Guided Proposal Network (IGPN), which can learn the similarity between query persons and proposals. Thus, we can decrease proposals according to the similarity scores. To incorporate information of the query into the detection network, we introduce the Siamese region proposal network to Faster-RCNN and we propose improved cross-correlation layers to alleviate the imbalance of parameters distribution. Furthermore, we design a local relation block and a global relation branch to leverage the proposal-proposal relations and query-scene relations, respectively. Extensive experiments show that our method improves the person search performance through decreasing proposals and achieves competitive performance on two large person search benchmark datasets, CUHK-SYSU and PRW.
{"title":"Instance Guided Proposal Network for Person Search","authors":"Wenkai Dong, Zhaoxiang Zhang, Chunfeng Song, T. Tan","doi":"10.1109/CVPR42600.2020.00266","DOIUrl":"https://doi.org/10.1109/CVPR42600.2020.00266","url":null,"abstract":"Person detection networks have been widely used in person search. These detectors discriminate persons from the background and generate proposals of all the persons from a gallery of scene images for each query. However, such a large number of proposals have a negative influence on the following identity matching process because many distractors are involved. In this paper, we propose a new detection network for person search, named Instance Guided Proposal Network (IGPN), which can learn the similarity between query persons and proposals. Thus, we can decrease proposals according to the similarity scores. To incorporate information of the query into the detection network, we introduce the Siamese region proposal network to Faster-RCNN and we propose improved cross-correlation layers to alleviate the imbalance of parameters distribution. Furthermore, we design a local relation block and a global relation branch to leverage the proposal-proposal relations and query-scene relations, respectively. Extensive experiments show that our method improves the person search performance through decreasing proposals and achieves competitive performance on two large person search benchmark datasets, CUHK-SYSU and PRW.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"487 1","pages":"2582-2591"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78829340","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}
In this work, we study various existing benchmarks for deepfake detection researches. In particular, we examine a novel two-stage face swapping algorithm, called FaceShifter, for high fidelity and occlusion aware face swapping. Unlike many existing face swapping works that leverage only limited information from the target image when synthesizing the swapped face, FaceShifter generates the swapped face with high-fidelity by exploiting and integrating the target attributes thoroughly and adaptively. FaceShifter can handle facial occlusions with a second synthesis stage consisting of a Heuristic Error Acknowledging Refinement Network (HEAR-Net), which is trained to recover anomaly regions in a self-supervised way without any manual annotations. Experiments show that existing deepfake detection algorithm performs poorly with FaceShifter, since it achieves advantageous quality over all existing benchmarks. However, our newly developed Face X-Ray method can reliably detect forged images created by FaceShifter.
{"title":"Advancing High Fidelity Identity Swapping for Forgery Detection","authors":"Lingzhi Li, Jianmin Bao, Hao Yang, Dong Chen, Fang Wen","doi":"10.1109/cvpr42600.2020.00512","DOIUrl":"https://doi.org/10.1109/cvpr42600.2020.00512","url":null,"abstract":"In this work, we study various existing benchmarks for deepfake detection researches. In particular, we examine a novel two-stage face swapping algorithm, called FaceShifter, for high fidelity and occlusion aware face swapping. Unlike many existing face swapping works that leverage only limited information from the target image when synthesizing the swapped face, FaceShifter generates the swapped face with high-fidelity by exploiting and integrating the target attributes thoroughly and adaptively. FaceShifter can handle facial occlusions with a second synthesis stage consisting of a Heuristic Error Acknowledging Refinement Network (HEAR-Net), which is trained to recover anomaly regions in a self-supervised way without any manual annotations. Experiments show that existing deepfake detection algorithm performs poorly with FaceShifter, since it achieves advantageous quality over all existing benchmarks. However, our newly developed Face X-Ray method can reliably detect forged images created by FaceShifter.","PeriodicalId":6715,"journal":{"name":"2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"59 1","pages":"5073-5082"},"PeriodicalIF":0.0,"publicationDate":"2020-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90824615","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：