Pub Date : 2021-06-01DOI: 10.1109/CVPR46437.2021.00278
A. Lahiri, Vivek Kwatra, C. Frueh, John Lewis, C. Bregler
In this paper, we present a video-based learning framework for animating personalized 3D talking faces from audio. We introduce two training-time data normalizations that significantly improve data sample efficiency. First, we isolate and represent faces in a normalized space that decouples 3D geometry, head pose, and texture. This decomposes the prediction problem into regressions over the 3D face shape and the corresponding 2D texture atlas. Second, we leverage facial symmetry and approximate albedo constancy of skin to isolate and remove spatio-temporal lighting variations. Together, these normalizations allow simple networks to generate high fidelity lip-sync videos under novel ambient illumination while training with just a single speaker-specific video. Further, to stabilize temporal dynamics, we introduce an auto-regressive approach that conditions the model on its previous visual state. Human ratings and objective metrics demonstrate that our method outperforms contemporary state-of-the-art audio-driven video reenactment benchmarks in terms of realism, lip-sync and visual quality scores. We illustrate several applications enabled by our framework.
{"title":"LipSync3D: Data-Efficient Learning of Personalized 3D Talking Faces from Video using Pose and Lighting Normalization","authors":"A. Lahiri, Vivek Kwatra, C. Frueh, John Lewis, C. Bregler","doi":"10.1109/CVPR46437.2021.00278","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00278","url":null,"abstract":"In this paper, we present a video-based learning framework for animating personalized 3D talking faces from audio. We introduce two training-time data normalizations that significantly improve data sample efficiency. First, we isolate and represent faces in a normalized space that decouples 3D geometry, head pose, and texture. This decomposes the prediction problem into regressions over the 3D face shape and the corresponding 2D texture atlas. Second, we leverage facial symmetry and approximate albedo constancy of skin to isolate and remove spatio-temporal lighting variations. Together, these normalizations allow simple networks to generate high fidelity lip-sync videos under novel ambient illumination while training with just a single speaker-specific video. Further, to stabilize temporal dynamics, we introduce an auto-regressive approach that conditions the model on its previous visual state. Human ratings and objective metrics demonstrate that our method outperforms contemporary state-of-the-art audio-driven video reenactment benchmarks in terms of realism, lip-sync and visual quality scores. We illustrate several applications enabled by our framework.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116889475","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00987
J. Xiong, Jian Wang, W. Heidrich, S. Nayar
We propose a new flash technique for low-light imaging, using deep-red light as an illuminating source. Our main observation is that in a dim environment, the human eye mainly uses rods for the perception of light, which are not sensitive to wavelengths longer than 620nm, yet the camera sensor still has a spectral response. We propose a novel modulation strategy when training a modern CNN model for guided image filtering, fusing a noisy RGB frame and a flash frame. This fusion network is further extended for video reconstruction. We have built a prototype with minor hardware adjustments and tested the new flash technique on a variety of static and dynamic scenes. The experimental results demonstrate that our method produces compelling reconstructions, even in extra dim conditions.
{"title":"Seeing in Extra Darkness Using a Deep-Red Flash","authors":"J. Xiong, Jian Wang, W. Heidrich, S. Nayar","doi":"10.1109/CVPR46437.2021.00987","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00987","url":null,"abstract":"We propose a new flash technique for low-light imaging, using deep-red light as an illuminating source. Our main observation is that in a dim environment, the human eye mainly uses rods for the perception of light, which are not sensitive to wavelengths longer than 620nm, yet the camera sensor still has a spectral response. We propose a novel modulation strategy when training a modern CNN model for guided image filtering, fusing a noisy RGB frame and a flash frame. This fusion network is further extended for video reconstruction. We have built a prototype with minor hardware adjustments and tested the new flash technique on a variety of static and dynamic scenes. The experimental results demonstrate that our method produces compelling reconstructions, even in extra dim conditions.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"265 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121116988","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00528
Zhiqiang Fu, Yao Zhao, Dongxia Chang, Xingxing Zhang, Yiming Wang
This paper presents a novel, simple yet robust self-representation method, i.e., Double Low-Rank Representation with Projection Distance penalty (DLRRPD) for clustering. With the learned optimal projected representations, DLRRPD is capable of obtaining an effective similarity graph to capture the multi-subspace structure. Besides the global low-rank constraint, the local geometrical structure is additionally exploited via a projection distance penalty in our DLRRPD, thus facilitating a more favorable graph. Moreover, to improve the robustness of DLRRPD to noises, we introduce a Laplacian rank constraint, which can further encourage the learned graph to be more discriminative for clustering tasks. Meanwhile, Frobenius norm (instead of the popularly used nuclear norm) is employed to enforce the graph to be more block-diagonal with lower complexity. Extensive experiments have been conducted on synthetic, real, and noisy data to show that the proposed method outperforms currently available alternatives by a margin of 1.0%~10.1%.
{"title":"Double low-rank representation with projection distance penalty for clustering","authors":"Zhiqiang Fu, Yao Zhao, Dongxia Chang, Xingxing Zhang, Yiming Wang","doi":"10.1109/CVPR46437.2021.00528","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00528","url":null,"abstract":"This paper presents a novel, simple yet robust self-representation method, i.e., Double Low-Rank Representation with Projection Distance penalty (DLRRPD) for clustering. With the learned optimal projected representations, DLRRPD is capable of obtaining an effective similarity graph to capture the multi-subspace structure. Besides the global low-rank constraint, the local geometrical structure is additionally exploited via a projection distance penalty in our DLRRPD, thus facilitating a more favorable graph. Moreover, to improve the robustness of DLRRPD to noises, we introduce a Laplacian rank constraint, which can further encourage the learned graph to be more discriminative for clustering tasks. Meanwhile, Frobenius norm (instead of the popularly used nuclear norm) is employed to enforce the graph to be more block-diagonal with lower complexity. Extensive experiments have been conducted on synthetic, real, and noisy data to show that the proposed method outperforms currently available alternatives by a margin of 1.0%~10.1%.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124814976","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00317
Shuo Yang, Min Xu, Haozhe Xie, Stuart W. Perry, Jiahao Xia
Existing methods for single-view 3D object reconstruction directly learn to transform image features into 3D representations. However, these methods are vulnerable to images containing noisy backgrounds and heavy occlusions because the extracted image features do not contain enough information to reconstruct high-quality 3D shapes. Humans routinely use incomplete or noisy visual cues from an image to retrieve similar 3D shapes from their memory and reconstruct the 3D shape of an object. Inspired by this, we propose a novel method, named Mem3D, that explicitly constructs shape priors to supplement the missing information in the image. Specifically, the shape priors are in the forms of "image-voxel" pairs in the memory network, which is stored by a well-designed writing strategy during training. We also propose a voxel triplet loss function that helps to retrieve the precise 3D shapes that are highly related to the input image from shape priors. The LSTM-based shape encoder is introduced to extract information from the retrieved 3D shapes, which are useful in recovering the 3D shape of an object that is heavily occluded or in complex environments. Experimental results demonstrate that Mem3D significantly improves reconstruction quality and performs favorably against state-of-the-art methods on the ShapeNet and Pix3D datasets.
{"title":"Single-View 3D Object Reconstruction from Shape Priors in Memory","authors":"Shuo Yang, Min Xu, Haozhe Xie, Stuart W. Perry, Jiahao Xia","doi":"10.1109/CVPR46437.2021.00317","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00317","url":null,"abstract":"Existing methods for single-view 3D object reconstruction directly learn to transform image features into 3D representations. However, these methods are vulnerable to images containing noisy backgrounds and heavy occlusions because the extracted image features do not contain enough information to reconstruct high-quality 3D shapes. Humans routinely use incomplete or noisy visual cues from an image to retrieve similar 3D shapes from their memory and reconstruct the 3D shape of an object. Inspired by this, we propose a novel method, named Mem3D, that explicitly constructs shape priors to supplement the missing information in the image. Specifically, the shape priors are in the forms of \"image-voxel\" pairs in the memory network, which is stored by a well-designed writing strategy during training. We also propose a voxel triplet loss function that helps to retrieve the precise 3D shapes that are highly related to the input image from shape priors. The LSTM-based shape encoder is introduced to extract information from the retrieved 3D shapes, which are useful in recovering the 3D shape of an object that is heavily occluded or in complex environments. Experimental results demonstrate that Mem3D significantly improves reconstruction quality and performs favorably against state-of-the-art methods on the ShapeNet and Pix3D datasets.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124827789","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00344
Xiao Zhang, Yixiao Ge, Y. Qiao, Hongsheng Li
Unsupervised object re-identification targets at learning discriminative representations for object retrieval without any annotations. Clustering-based methods [27], [46], [10] conduct training with the generated pseudo labels and currently dominate this research direction. However, they still suffer from the issue of pseudo label noise. To tackle the challenge, we propose to properly estimate pseudo label similarities between consecutive training generations with clustering consensus and refine pseudo labels with temporally propagated and ensembled pseudo labels. To the best of our knowledge, this is the first attempt to leverage the spirit of temporal ensembling [25] to improve classification with dynamically changing classes over generations. The proposed pseudo label refinery strategy is simple yet effective and can be seamlessly integrated into existing clustering-based unsupervised re-identification methods. With our proposed approach, state-of-the-art method [10] can be further boosted with up to 8.8% mAP improvements on the challenging MSMT17 [39] dataset.
{"title":"Refining Pseudo Labels with Clustering Consensus over Generations for Unsupervised Object Re-identification","authors":"Xiao Zhang, Yixiao Ge, Y. Qiao, Hongsheng Li","doi":"10.1109/CVPR46437.2021.00344","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00344","url":null,"abstract":"Unsupervised object re-identification targets at learning discriminative representations for object retrieval without any annotations. Clustering-based methods [27], [46], [10] conduct training with the generated pseudo labels and currently dominate this research direction. However, they still suffer from the issue of pseudo label noise. To tackle the challenge, we propose to properly estimate pseudo label similarities between consecutive training generations with clustering consensus and refine pseudo labels with temporally propagated and ensembled pseudo labels. To the best of our knowledge, this is the first attempt to leverage the spirit of temporal ensembling [25] to improve classification with dynamically changing classes over generations. The proposed pseudo label refinery strategy is simple yet effective and can be seamlessly integrated into existing clustering-based unsupervised re-identification methods. With our proposed approach, state-of-the-art method [10] can be further boosted with up to 8.8% mAP improvements on the challenging MSMT17 [39] dataset.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123689815","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}
Monocular 3D object detection is an important task in autonomous driving. It can be easily intractable where there exists ego-car pose change w.r.t. ground plane. This is common due to the slight fluctuation of road smoothness and slope. Due to the lack of insight in industrial application, existing methods on open datasets neglect the cam-era pose information, which inevitably results in the detector being susceptible to camera extrinsic parameters. The perturbation of objects is very popular in most autonomous driving cases for industrial products. To this end, we propose a novel method to capture camera pose to formulate the detector free from extrinsic perturbation. Specifically, the proposed framework predicts camera extrinsic parameters by detecting vanishing point and horizon change. A converter is designed to rectify perturbative features in the latent space. By doing so, our 3D detector works independent of the extrinsic parameter variations and produces accurate results in realistic cases, e.g., potholed and uneven roads, where almost all existing monocular detectors fail to handle. Experiments demonstrate our method yields the best performance compared with the other state-of-the-arts by a large margin on both KITTI 3D and nuScenes datasets.
{"title":"Monocular 3D Object Detection: An Extrinsic Parameter Free Approach","authors":"Yunsong Zhou, Yuan He, Hongzi Zhu, Cheng Wang, Hongyang Li, Qinhong Jiang","doi":"10.1109/CVPR46437.2021.00747","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00747","url":null,"abstract":"Monocular 3D object detection is an important task in autonomous driving. It can be easily intractable where there exists ego-car pose change w.r.t. ground plane. This is common due to the slight fluctuation of road smoothness and slope. Due to the lack of insight in industrial application, existing methods on open datasets neglect the cam-era pose information, which inevitably results in the detector being susceptible to camera extrinsic parameters. The perturbation of objects is very popular in most autonomous driving cases for industrial products. To this end, we propose a novel method to capture camera pose to formulate the detector free from extrinsic perturbation. Specifically, the proposed framework predicts camera extrinsic parameters by detecting vanishing point and horizon change. A converter is designed to rectify perturbative features in the latent space. By doing so, our 3D detector works independent of the extrinsic parameter variations and produces accurate results in realistic cases, e.g., potholed and uneven roads, where almost all existing monocular detectors fail to handle. Experiments demonstrate our method yields the best performance compared with the other state-of-the-arts by a large margin on both KITTI 3D and nuScenes datasets.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126852387","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00335
Karren D. Yang, Wan-Yi Lin, M. Barman, Filipe Condessa, Zico Kolter
Beyond achieving high performance across many vision tasks, multimodal models are expected to be robust to single-source faults due to the availability of redundant information between modalities. In this paper, we investigate the robustness of multimodal neural networks against worst-case (i.e., adversarial) perturbations on a single modality. We first show that standard multimodal fusion models are vulnerable to single-source adversaries: an attack on any single modality can overcome the correct information from multiple unperturbed modalities and cause the model to fail. This surprising vulnerability holds across diverse multimodal tasks and necessitates a solution. Motivated by this finding, we propose an adversarially robust fusion strategy that trains the model to compare information coming from all the input sources, detect inconsistencies in the perturbed modality compared to the other modalities, and only allow information from the unperturbed modalities to pass through. Our approach significantly improves on state-of-the-art methods in single-source robustness, achieving gains of 7.8-25.2% on action recognition, 19.7-48.2% on object detection, and 1.6-6.7% on sentiment analysis, without degrading performance on unperturbed (i.e., clean) data.
{"title":"Defending Multimodal Fusion Models against Single-Source Adversaries","authors":"Karren D. Yang, Wan-Yi Lin, M. Barman, Filipe Condessa, Zico Kolter","doi":"10.1109/CVPR46437.2021.00335","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00335","url":null,"abstract":"Beyond achieving high performance across many vision tasks, multimodal models are expected to be robust to single-source faults due to the availability of redundant information between modalities. In this paper, we investigate the robustness of multimodal neural networks against worst-case (i.e., adversarial) perturbations on a single modality. We first show that standard multimodal fusion models are vulnerable to single-source adversaries: an attack on any single modality can overcome the correct information from multiple unperturbed modalities and cause the model to fail. This surprising vulnerability holds across diverse multimodal tasks and necessitates a solution. Motivated by this finding, we propose an adversarially robust fusion strategy that trains the model to compare information coming from all the input sources, detect inconsistencies in the perturbed modality compared to the other modalities, and only allow information from the unperturbed modalities to pass through. Our approach significantly improves on state-of-the-art methods in single-source robustness, achieving gains of 7.8-25.2% on action recognition, 19.7-48.2% on object detection, and 1.6-6.7% on sentiment analysis, without degrading performance on unperturbed (i.e., clean) data.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114939049","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.01566
Olivia Wiles, Sébastien Ehrhardt, Andrew Zisserman
We propose a new approach to determine correspondences between image pairs in the wild under large changes in illumination, viewpoint, context, and material. While other approaches find correspondences between pairs of images by treating the images independently, we instead condition on both images to implicitly take account of the differences between them. To achieve this, we introduce (i) a spatial attention mechanism (a co-attention module, CoAM) for conditioning the learned features on both images, and (ii) a distinctiveness score used to choose the best matches at test time. CoAM can be added to standard architectures and trained using self-supervision or supervised data, and achieves a significant performance improvement under hard conditions, e.g. large viewpoint changes. We demonstrate that models using CoAM achieve state of the art or competitive results on a wide range of tasks: local matching, camera localization, 3D reconstruction, and image stylization.
{"title":"Co-Attention for Conditioned Image Matching","authors":"Olivia Wiles, Sébastien Ehrhardt, Andrew Zisserman","doi":"10.1109/CVPR46437.2021.01566","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.01566","url":null,"abstract":"We propose a new approach to determine correspondences between image pairs in the wild under large changes in illumination, viewpoint, context, and material. While other approaches find correspondences between pairs of images by treating the images independently, we instead condition on both images to implicitly take account of the differences between them. To achieve this, we introduce (i) a spatial attention mechanism (a co-attention module, CoAM) for conditioning the learned features on both images, and (ii) a distinctiveness score used to choose the best matches at test time. CoAM can be added to standard architectures and trained using self-supervision or supervised data, and achieves a significant performance improvement under hard conditions, e.g. large viewpoint changes. We demonstrate that models using CoAM achieve state of the art or competitive results on a wide range of tasks: local matching, camera localization, 3D reconstruction, and image stylization.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"261 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116032798","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}
Though machine learning algorithms are able to achieve pattern recognition from the correlation between data and labels, the presence of spurious features in the data decreases the robustness of these learned relationships with respect to varied testing environments. This is known as out-of-distribution (OoD) generalization problem. Recently, invariant risk minimization (IRM) attempts to tackle this issue by penalizing predictions based on the unstable spurious features in the data collected from different environments. However, similar to domain adaptation or domain generalization, a prevalent non-trivial limitation in these works is that the environment information is assigned by human specialists, i.e. a priori, or determined heuristically. However, an inappropriate group partitioning can dramatically deteriorate the OoD generalization and this process is expensive and time-consuming. To deal with this issue, we propose a novel theoretically principled min-max framework to iteratively construct a worst-case splitting, i.e. creating the most challenging environment splittings for the backbone learning paradigm (e.g. IRM) to learn the robust feature representation. We also design a differentiable training strategy to facilitate the feasible gradient- based computation. Numerical experiments show that our algorithmic framework has achieved superior and stable performance in various datasets, such as Colored MNIST and Punctuated Stanford sentiment treebank (SST). Furthermore, we also find our algorithm to be robust even to a strong data poisoning attack. To the best of our knowledge, this is one of the first to adopt differentiable environment splitting method to enable stable predictions across environments without environment index information, which achieves the state-of-the-art performance on datasets with strong spurious correlation, such as Colored MNIST.
{"title":"Adversarial Invariant Learning","authors":"Nanyang Ye, Jingxuan Tang, Huayu Deng, Xiao-Yun Zhou, Qianxiao Li, Zhenguo Li, Guang-Zhong Yang, Zhanxing Zhu","doi":"10.1109/CVPR46437.2021.01226","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.01226","url":null,"abstract":"Though machine learning algorithms are able to achieve pattern recognition from the correlation between data and labels, the presence of spurious features in the data decreases the robustness of these learned relationships with respect to varied testing environments. This is known as out-of-distribution (OoD) generalization problem. Recently, invariant risk minimization (IRM) attempts to tackle this issue by penalizing predictions based on the unstable spurious features in the data collected from different environments. However, similar to domain adaptation or domain generalization, a prevalent non-trivial limitation in these works is that the environment information is assigned by human specialists, i.e. a priori, or determined heuristically. However, an inappropriate group partitioning can dramatically deteriorate the OoD generalization and this process is expensive and time-consuming. To deal with this issue, we propose a novel theoretically principled min-max framework to iteratively construct a worst-case splitting, i.e. creating the most challenging environment splittings for the backbone learning paradigm (e.g. IRM) to learn the robust feature representation. We also design a differentiable training strategy to facilitate the feasible gradient- based computation. Numerical experiments show that our algorithmic framework has achieved superior and stable performance in various datasets, such as Colored MNIST and Punctuated Stanford sentiment treebank (SST). Furthermore, we also find our algorithm to be robust even to a strong data poisoning attack. To the best of our knowledge, this is one of the first to adopt differentiable environment splitting method to enable stable predictions across environments without environment index information, which achieves the state-of-the-art performance on datasets with strong spurious correlation, such as Colored MNIST.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"162 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116043458","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 : 2021-06-01DOI: 10.1109/CVPR46437.2021.00246
Zhanyu Wang, Luping Zhou, Lei Wang, Xiu Li
Automated radiographic report generation is a challenging task since it requires to generate paragraphs describing fine-grained visual differences of cases, especially for those between the diseased and the healthy. Existing image captioning methods commonly target at generic images, and lack mechanism to meet this requirement. To bridge this gap, in this paper, we propose a self-boosting framework that improves radiographic report generation based on the cooperation of the main task of report generation and an auxiliary task of image-text matching. The two tasks are built as the two branches of a network model and influence each other in a cooperative way. On one hand, the image-text matching branch helps to learn highly text-correlated visual features for the report generation branch to output high quality reports. On the other hand, the improved reports produced by the report generation branch provide additional harder samples for the image-text matching branch and enforce the latter to improve itself by learning better visual and text feature representations. This, in turn, helps improve the report generation branch again. These two branches are jointly trained to help improve each other iteratively and progressively, so that the whole model is self-boosted without requiring external resources. Experimental results demonstrate the effectiveness of our method on two public datasets, showing its superior performance over multiple state-of-the-art image captioning and medical report generation methods.
{"title":"A Self-boosting Framework for Automated Radiographic Report Generation","authors":"Zhanyu Wang, Luping Zhou, Lei Wang, Xiu Li","doi":"10.1109/CVPR46437.2021.00246","DOIUrl":"https://doi.org/10.1109/CVPR46437.2021.00246","url":null,"abstract":"Automated radiographic report generation is a challenging task since it requires to generate paragraphs describing fine-grained visual differences of cases, especially for those between the diseased and the healthy. Existing image captioning methods commonly target at generic images, and lack mechanism to meet this requirement. To bridge this gap, in this paper, we propose a self-boosting framework that improves radiographic report generation based on the cooperation of the main task of report generation and an auxiliary task of image-text matching. The two tasks are built as the two branches of a network model and influence each other in a cooperative way. On one hand, the image-text matching branch helps to learn highly text-correlated visual features for the report generation branch to output high quality reports. On the other hand, the improved reports produced by the report generation branch provide additional harder samples for the image-text matching branch and enforce the latter to improve itself by learning better visual and text feature representations. This, in turn, helps improve the report generation branch again. These two branches are jointly trained to help improve each other iteratively and progressively, so that the whole model is self-boosted without requiring external resources. Experimental results demonstrate the effectiveness of our method on two public datasets, showing its superior performance over multiple state-of-the-art image captioning and medical report generation methods.","PeriodicalId":339646,"journal":{"name":"2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","volume":"135 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122402282","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：