Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.00376
Huajun Liu, Xiangyu Miao, C. Mertz, Chengzhong Xu, Hui Kong
Cracks are irregular line structures that are of interest in many computer vision applications. Crack detection (e.g., from pavement images) is a challenging task due to intensity in-homogeneity, topology complexity, low contrast and noisy background. The overall crack detection accuracy can be significantly affected by the detection performance on fine-grained cracks. In this work, we propose a Crack Transformer network (CrackFormer) for fine-grained crack detection. The CrackFormer is composed of novel attention modules in a SegNet-like encoder-decoder architecture. Specifically, it consists of novel self-attention modules with 1x1 convolutional kernels for efficient contextual information extraction across feature-channels, and efficient positional embedding to capture large receptive field contextual information for long range interactions. It also introduces new scaling-attention modules to combine outputs from the corresponding encoder and decoder blocks to suppress non-semantic features and sharpen semantic ones. The CrackFormer is trained and evaluated on three classical crack datasets. The experimental results show that the CrackFormer achieves the Optimal Dataset Scale (ODS) values of 0.871, 0.877 and 0.881, respectively, on the three datasets and outperforms the state-of-the-art methods.
{"title":"CrackFormer: Transformer Network for Fine-Grained Crack Detection","authors":"Huajun Liu, Xiangyu Miao, C. Mertz, Chengzhong Xu, Hui Kong","doi":"10.1109/ICCV48922.2021.00376","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00376","url":null,"abstract":"Cracks are irregular line structures that are of interest in many computer vision applications. Crack detection (e.g., from pavement images) is a challenging task due to intensity in-homogeneity, topology complexity, low contrast and noisy background. The overall crack detection accuracy can be significantly affected by the detection performance on fine-grained cracks. In this work, we propose a Crack Transformer network (CrackFormer) for fine-grained crack detection. The CrackFormer is composed of novel attention modules in a SegNet-like encoder-decoder architecture. Specifically, it consists of novel self-attention modules with 1x1 convolutional kernels for efficient contextual information extraction across feature-channels, and efficient positional embedding to capture large receptive field contextual information for long range interactions. It also introduces new scaling-attention modules to combine outputs from the corresponding encoder and decoder blocks to suppress non-semantic features and sharpen semantic ones. The CrackFormer is trained and evaluated on three classical crack datasets. The experimental results show that the CrackFormer achieves the Optimal Dataset Scale (ODS) values of 0.871, 0.877 and 0.881, respectively, on the three datasets and outperforms the state-of-the-art methods.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"5 1","pages":"3763-3772"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74316604","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.00533
Qi Sun, Chen Bai, Tinghuan Chen, Hao Geng, Xinyun Zhang, Yang Bai, Bei Yu
Deep neural networks (DNNs) have been widely used recently while their hardware deployment optimizations are very time-consuming and the historical deployment knowledge is not utilized efficiently. In this paper, to accelerate the optimization process and find better deployment configurations, we propose a novel transfer learning method based on deep Gaussian processes (DGPs). Firstly, a deep Gaussian process (DGP) model is built on the historical data to learn empirical knowledge. Secondly, to transfer knowledge to a new task, a tuning set is sampled for the new task under the guidance of the DGP model. Then DGP is tuned according to the tuning set via maximum-a-posteriori (MAP) estimation to accommodate for the new task and finally used to guide the deployments of the task. The experiments show that our method achieves the best inference latencies of convolutions while accelerating the optimization process significantly, compared with previous arts.
{"title":"Fast and Efficient DNN Deployment via Deep Gaussian Transfer Learning","authors":"Qi Sun, Chen Bai, Tinghuan Chen, Hao Geng, Xinyun Zhang, Yang Bai, Bei Yu","doi":"10.1109/ICCV48922.2021.00533","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00533","url":null,"abstract":"Deep neural networks (DNNs) have been widely used recently while their hardware deployment optimizations are very time-consuming and the historical deployment knowledge is not utilized efficiently. In this paper, to accelerate the optimization process and find better deployment configurations, we propose a novel transfer learning method based on deep Gaussian processes (DGPs). Firstly, a deep Gaussian process (DGP) model is built on the historical data to learn empirical knowledge. Secondly, to transfer knowledge to a new task, a tuning set is sampled for the new task under the guidance of the DGP model. Then DGP is tuned according to the tuning set via maximum-a-posteriori (MAP) estimation to accommodate for the new task and finally used to guide the deployments of the task. The experiments show that our method achieves the best inference latencies of convolutions while accelerating the optimization process significantly, compared with previous arts.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"27 1","pages":"5360-5370"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75647267","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 : 2021-10-01DOI: 10.1109/iccv48922.2021.00234
Aviad Levis, Daeyoung Lee, J. Tropp, C. Gammie, K. Bouman
We develop an approach to recover the underlying properties of fluid-dynamical processes from sparse measurements. We are motivated by the task of imaging the stochastically evolving environment surrounding black holes, and demonstrate how flow parameters can be estimated from sparse interferometric measurements used in radio astronomical imaging. To model the stochastic flow we use spatio-temporal Gaussian Random Fields (GRFs). The high dimensionality of the underlying source video makes direct representation via a GRF’s full covariance matrix intractable. In contrast, stochastic partial differential equations are able to capture correlations at multiple scales by specifying only local interaction coefficients. Our approach estimates the coefficients of a space-time diffusion equation that dictates the stationary statistics of the dynamical process. We analyze our approach on realistic simulations of black hole evolution and demonstrate its advantage over state-of-the-art dynamic black hole imaging techniques.
{"title":"Inference of Black Hole Fluid-Dynamics from Sparse Interferometric Measurements","authors":"Aviad Levis, Daeyoung Lee, J. Tropp, C. Gammie, K. Bouman","doi":"10.1109/iccv48922.2021.00234","DOIUrl":"https://doi.org/10.1109/iccv48922.2021.00234","url":null,"abstract":"We develop an approach to recover the underlying properties of fluid-dynamical processes from sparse measurements. We are motivated by the task of imaging the stochastically evolving environment surrounding black holes, and demonstrate how flow parameters can be estimated from sparse interferometric measurements used in radio astronomical imaging. To model the stochastic flow we use spatio-temporal Gaussian Random Fields (GRFs). The high dimensionality of the underlying source video makes direct representation via a GRF’s full covariance matrix intractable. In contrast, stochastic partial differential equations are able to capture correlations at multiple scales by specifying only local interaction coefficients. Our approach estimates the coefficients of a space-time diffusion equation that dictates the stationary statistics of the dynamical process. We analyze our approach on realistic simulations of black hole evolution and demonstrate its advantage over state-of-the-art dynamic black hole imaging techniques.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"74 1","pages":"2320-2329"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74445243","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.01566
Chenxi Wang, Haoshu Fang, Minghao Gou, Hongjie Fang, Jin Gao, Cewu Lu, S. Tong
Efficient and robust grasp pose detection is vital for robotic manipulation. For general 6 DoF grasping, conventional methods treat all points in a scene equally and usually adopt uniform sampling to select grasp candidates. However, we discover that ignoring where to grasp greatly harms the speed and accuracy of current grasp pose detection methods. In this paper, we propose "graspness", a quality based on geometry cues that distinguishes graspable area in cluttered scenes. A look-ahead searching method is proposed for measuring the graspness and statistical results justify the rationality of our method. To quickly detect graspness in practice, we develop a neural network named graspness model to approximate the searching process. Extensive experiments verify the stability, generality and effectiveness of our graspness model, allowing it to be used as a plug-and-play module for different methods. A large improvement in accuracy is witnessed for various previous methods after equipping our graspness model. Moreover, we develop GSNet, an end-to-end network that incorporate our graspness model for early filtering of low quality predictions. Experiments on a large scale benchmark, GraspNet-1Billion, show that our method outperforms previous arts by a large margin (30 + AP) and achieves a high inference speed.
{"title":"Graspness Discovery in Clutters for Fast and Accurate Grasp Detection","authors":"Chenxi Wang, Haoshu Fang, Minghao Gou, Hongjie Fang, Jin Gao, Cewu Lu, S. Tong","doi":"10.1109/ICCV48922.2021.01566","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01566","url":null,"abstract":"Efficient and robust grasp pose detection is vital for robotic manipulation. For general 6 DoF grasping, conventional methods treat all points in a scene equally and usually adopt uniform sampling to select grasp candidates. However, we discover that ignoring where to grasp greatly harms the speed and accuracy of current grasp pose detection methods. In this paper, we propose \"graspness\", a quality based on geometry cues that distinguishes graspable area in cluttered scenes. A look-ahead searching method is proposed for measuring the graspness and statistical results justify the rationality of our method. To quickly detect graspness in practice, we develop a neural network named graspness model to approximate the searching process. Extensive experiments verify the stability, generality and effectiveness of our graspness model, allowing it to be used as a plug-and-play module for different methods. A large improvement in accuracy is witnessed for various previous methods after equipping our graspness model. Moreover, we develop GSNet, an end-to-end network that incorporate our graspness model for early filtering of low quality predictions. Experiments on a large scale benchmark, GraspNet-1Billion, show that our method outperforms previous arts by a large margin (30 + AP) and achieves a high inference speed.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"107 1","pages":"15944-15953"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75696066","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.00622
Na Lei, X. Gu
An optimal transportation map finds the most economical way to transport one probability measure to the other. It has been applied in a broad range of applications in vision, deep learning and medical images. By Brenier theory, computing the optimal transport map is equivalent to solving a Monge-Ampère equation. Due to the highly non-linear nature, the computation of optimal transportation maps in large scale is very challenging.This work proposes a simple but powerful method, the FFT-OT algorithm, to tackle this difficulty based on three key ideas. First, solving Monge-Ampère equation is converted to a fixed point problem; Second, the obliqueness property of optimal transportation maps are reformulated as Neumann boundary conditions on rectangular domains; Third, FFT is applied in each iteration to solve a Poisson equation in order to improve the efficiency.Experiments on surfaces captured from 3D scanning and reconstructed from medical imaging are conducted, and compared with other existing methods. Our experimental results show that the proposed FFT-OT algorithm is simple, general and scalable with high efficiency and accuracy.
{"title":"FFT-OT: A Fast Algorithm for Optimal Transportation","authors":"Na Lei, X. Gu","doi":"10.1109/ICCV48922.2021.00622","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00622","url":null,"abstract":"An optimal transportation map finds the most economical way to transport one probability measure to the other. It has been applied in a broad range of applications in vision, deep learning and medical images. By Brenier theory, computing the optimal transport map is equivalent to solving a Monge-Ampère equation. Due to the highly non-linear nature, the computation of optimal transportation maps in large scale is very challenging.This work proposes a simple but powerful method, the FFT-OT algorithm, to tackle this difficulty based on three key ideas. First, solving Monge-Ampère equation is converted to a fixed point problem; Second, the obliqueness property of optimal transportation maps are reformulated as Neumann boundary conditions on rectangular domains; Third, FFT is applied in each iteration to solve a Poisson equation in order to improve the efficiency.Experiments on surfaces captured from 3D scanning and reconstructed from medical imaging are conducted, and compared with other existing methods. Our experimental results show that the proposed FFT-OT algorithm is simple, general and scalable with high efficiency and accuracy.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"106 1","pages":"6260-6269"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75741546","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.01176
Songtao He, M. Sadeghi, S. Chawla, Mohammad Alizadeh, Harinarayanan Balakrishnan, Sam Madden
Traffic accidents cost about 3% of the world’s GDP and are the leading cause of death in children and young adults. Accident risk maps are useful tools to monitor and mitigate accident risk. We present a technique to generate high-resolution (5 meters) accident risk maps. At this high resolution, accidents are sparse and risk estimation is limited by bias-variance trade-off. Prior accident risk maps either estimate low-resolution maps that are of low utility (high bias), or they use frequency-based estimation techniques that inaccurately predict where accidents actually happen (high variance). To improve this trade-off, we use an end-to-end deep architecture that can input satellite imagery, GPS trajectories, road maps and the history of accidents. Our evaluation on four metropolitan areas in the US with a total area of 7,488 km2 shows that our technique outperform prior work in terms of resolution and accuracy.
{"title":"Inferring high-resolution traffic accident risk maps based on satellite imagery and GPS trajectories","authors":"Songtao He, M. Sadeghi, S. Chawla, Mohammad Alizadeh, Harinarayanan Balakrishnan, Sam Madden","doi":"10.1109/ICCV48922.2021.01176","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01176","url":null,"abstract":"Traffic accidents cost about 3% of the world’s GDP and are the leading cause of death in children and young adults. Accident risk maps are useful tools to monitor and mitigate accident risk. We present a technique to generate high-resolution (5 meters) accident risk maps. At this high resolution, accidents are sparse and risk estimation is limited by bias-variance trade-off. Prior accident risk maps either estimate low-resolution maps that are of low utility (high bias), or they use frequency-based estimation techniques that inaccurately predict where accidents actually happen (high variance). To improve this trade-off, we use an end-to-end deep architecture that can input satellite imagery, GPS trajectories, road maps and the history of accidents. Our evaluation on four metropolitan areas in the US with a total area of 7,488 km2 shows that our technique outperform prior work in terms of resolution and accuracy.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"56 1","pages":"11957-11965"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73858121","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.00441
Zhuo Zheng, Wenqi Ren, Xiaochun Cao, Tao Wang, Xiuyi Jia
Existing single image high dynamic range (HDR) reconstruction methods attempt to expand the range of illuminance. They are not effective in generating plausible textures and colors in the reconstructed results, especially for high-density pixels in ultra-high-definition (UHD) images. To address these problems, we propose a new HDR reconstruction network for UHD images by collaboratively learning color and texture details. First, we propose a dual-path network to extract the content and chromatic features at a reduced resolution of the low dynamic range (LDR) input. These two types of features are used to fit bilateral-space affine models for real-time HDR reconstruction. To extract the main data structure of the LDR input, we propose to use 3D Tucker decomposition and reconstruction to prevent pseudo edges and noise amplification in the learned bilateral grid. As a result, the high-quality content and chromatic features can be reconstructed capitalized on guided bilateral upsampling. Finally, we fuse these two full-resolution feature maps into the HDR reconstructed results. Our proposed method can achieve real-time processing for UHD images (about 160 fps). Experimental results demonstrate that the proposed algorithm performs favorably against the state-of-the-art HDR reconstruction approaches on public benchmarks and real-world UHD images.
{"title":"Ultra-High-Definition Image HDR Reconstruction via Collaborative Bilateral Learning","authors":"Zhuo Zheng, Wenqi Ren, Xiaochun Cao, Tao Wang, Xiuyi Jia","doi":"10.1109/ICCV48922.2021.00441","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00441","url":null,"abstract":"Existing single image high dynamic range (HDR) reconstruction methods attempt to expand the range of illuminance. They are not effective in generating plausible textures and colors in the reconstructed results, especially for high-density pixels in ultra-high-definition (UHD) images. To address these problems, we propose a new HDR reconstruction network for UHD images by collaboratively learning color and texture details. First, we propose a dual-path network to extract the content and chromatic features at a reduced resolution of the low dynamic range (LDR) input. These two types of features are used to fit bilateral-space affine models for real-time HDR reconstruction. To extract the main data structure of the LDR input, we propose to use 3D Tucker decomposition and reconstruction to prevent pseudo edges and noise amplification in the learned bilateral grid. As a result, the high-quality content and chromatic features can be reconstructed capitalized on guided bilateral upsampling. Finally, we fuse these two full-resolution feature maps into the HDR reconstructed results. Our proposed method can achieve real-time processing for UHD images (about 160 fps). Experimental results demonstrate that the proposed algorithm performs favorably against the state-of-the-art HDR reconstruction approaches on public benchmarks and real-world UHD images.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"48 1","pages":"4429-4438"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80025591","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.01373
Jongyoo Kim, Jiaolong Yang, Xin Tong
For face texture completion, previous methods typically use some complete textures captured by multiview imaging systems or 3D scanners for supervised learning. This paper deals with a new challenging problem - learning to complete invisible texture in a single face image without using any complete texture. We simply leverage a large corpus of face images of different subjects (e. g., FFHQ) to train a texture completion model in an unsupervised manner. To achieve this, we propose DSD-GAN, a novel deep neural network based method that applies two discriminators in UV map space and image space. These two discriminators work in a complementary manner to learn both facial structures and texture details. We show that their combination is essential to obtain high-fidelity results. Despite the network never sees any complete facial appearance, it is able to generate compelling full textures from single images.
{"title":"Learning High-Fidelity Face Texture Completion without Complete Face Texture","authors":"Jongyoo Kim, Jiaolong Yang, Xin Tong","doi":"10.1109/ICCV48922.2021.01373","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01373","url":null,"abstract":"For face texture completion, previous methods typically use some complete textures captured by multiview imaging systems or 3D scanners for supervised learning. This paper deals with a new challenging problem - learning to complete invisible texture in a single face image without using any complete texture. We simply leverage a large corpus of face images of different subjects (e. g., FFHQ) to train a texture completion model in an unsupervised manner. To achieve this, we propose DSD-GAN, a novel deep neural network based method that applies two discriminators in UV map space and image space. These two discriminators work in a complementary manner to learn both facial structures and texture details. We show that their combination is essential to obtain high-fidelity results. Despite the network never sees any complete facial appearance, it is able to generate compelling full textures from single images.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"8 1","pages":"13970-13979"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80234233","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.01441
Ge Gao, P. You, Rong Pan, Shunyuan Han, Yuanyuan Zhang, Yuchao Dai, Ho-Jun Lee
In recent years, neural image compression emerges as a rapidly developing topic in computer vision, where the state-of-the-art approaches now exhibit superior compression performance than their conventional counterparts. Despite the great progress, current methods still have limitations in preserving fine spatial details for optimal reconstruction, especially at low compression rates. We make three contributions in tackling this issue. First, we develop a novel back projection method with attentional and multi-scale feature fusion for augmented representation power. Our back projection method recalibrates the current estimation by establishing feedback connections between high-level and low-level attributes in an attentional and discriminative manner. Second, we propose to decompose the input image and separately process the distinct frequency components, whose derived latents are recombined using a novel dual attention module, so that details inside regions of interest could be explicitly manipulated. Third, we propose a novel training scheme for reducing the latent rounding residual. Experimental results show that, when measured in PSNR, our model reduces BD-rate by 9.88% and 10.32% over the state-of-the-art method, and 4.12% and 4.32% over the latest coding standard Versatile Video Coding (VVC) on the Kodak and CLIC2020 Professional Validation dataset, respectively. Our approach also produces more visually pleasant images when optimized for MS-SSIM. The significant improvement upon existing methods shows the effectiveness of our method in preserving and remedying spatial information for enhanced compression quality.
{"title":"Neural Image Compression via Attentional Multi-scale Back Projection and Frequency Decomposition","authors":"Ge Gao, P. You, Rong Pan, Shunyuan Han, Yuanyuan Zhang, Yuchao Dai, Ho-Jun Lee","doi":"10.1109/ICCV48922.2021.01441","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01441","url":null,"abstract":"In recent years, neural image compression emerges as a rapidly developing topic in computer vision, where the state-of-the-art approaches now exhibit superior compression performance than their conventional counterparts. Despite the great progress, current methods still have limitations in preserving fine spatial details for optimal reconstruction, especially at low compression rates. We make three contributions in tackling this issue. First, we develop a novel back projection method with attentional and multi-scale feature fusion for augmented representation power. Our back projection method recalibrates the current estimation by establishing feedback connections between high-level and low-level attributes in an attentional and discriminative manner. Second, we propose to decompose the input image and separately process the distinct frequency components, whose derived latents are recombined using a novel dual attention module, so that details inside regions of interest could be explicitly manipulated. Third, we propose a novel training scheme for reducing the latent rounding residual. Experimental results show that, when measured in PSNR, our model reduces BD-rate by 9.88% and 10.32% over the state-of-the-art method, and 4.12% and 4.32% over the latest coding standard Versatile Video Coding (VVC) on the Kodak and CLIC2020 Professional Validation dataset, respectively. Our approach also produces more visually pleasant images when optimized for MS-SSIM. The significant improvement upon existing methods shows the effectiveness of our method in preserving and remedying spatial information for enhanced compression quality.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"28 1","pages":"14657-14666"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80293685","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 : 2021-10-01DOI: 10.1109/ICCV48922.2021.00961
Moitreya Chatterjee, N. Ahuja, A. Cherian
Predicting the future frames of a video is a challenging task, in part due to the underlying stochastic real-world phenomena. Prior approaches to solve this task typically estimate a latent prior characterizing this stochasticity, however do not account for the predictive uncertainty of the (deep learning) model. Such approaches often derive the training signal from the mean-squared error (MSE) between the generated frame and the ground truth, which can lead to sub-optimal training, especially when the predictive uncertainty is high. Towards this end, we introduce Neural Uncertainty Quantifier (NUQ) - a stochastic quantification of the model’s predictive uncertainty, and use it to weigh the MSE loss. We propose a hierarchical, variational framework to derive NUQ in a principled manner using a deep, Bayesian graphical model. Our experiments on three benchmark stochastic video prediction datasets show that our proposed framework trains more effectively compared to the state-of-the-art models (especially when the training sets are small), while demonstrating better video generation quality and diversity against several evaluation metrics.
{"title":"A Hierarchical Variational Neural Uncertainty Model for Stochastic Video Prediction","authors":"Moitreya Chatterjee, N. Ahuja, A. Cherian","doi":"10.1109/ICCV48922.2021.00961","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00961","url":null,"abstract":"Predicting the future frames of a video is a challenging task, in part due to the underlying stochastic real-world phenomena. Prior approaches to solve this task typically estimate a latent prior characterizing this stochasticity, however do not account for the predictive uncertainty of the (deep learning) model. Such approaches often derive the training signal from the mean-squared error (MSE) between the generated frame and the ground truth, which can lead to sub-optimal training, especially when the predictive uncertainty is high. Towards this end, we introduce Neural Uncertainty Quantifier (NUQ) - a stochastic quantification of the model’s predictive uncertainty, and use it to weigh the MSE loss. We propose a hierarchical, variational framework to derive NUQ in a principled manner using a deep, Bayesian graphical model. Our experiments on three benchmark stochastic video prediction datasets show that our proposed framework trains more effectively compared to the state-of-the-art models (especially when the training sets are small), while demonstrating better video generation quality and diversity against several evaluation metrics.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"125 1","pages":"9731-9741"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79022397","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：