Computer vision - ECCV 2022 . Part VII : 17th European Conference, Tel Aviv, Israel, October 23-27, 2022 : proceedings / / Shai Avidan [and four others]

• Pubblicazione: Cham, Switzerland : , : Springer, , [2022]; ©2022
• Descrizione fisica: 1 online resource (800 pages)
• Disciplina: 006.37
• Soggetto topico: Computer vision; Pattern recognition systems
• Persona (resp. second.): AvidanShai
• Nota di contenuto: Intro -- Foreword -- Preface -- Organization -- Contents - Part VII -- CT2: Colorization Transformer via Color Tokens -- 1 Introduction -- 2 Related Works -- 3 Method -- 3.1 Image Encoder -- 3.2 Color Encoder -- 3.3 Decoder -- 3.4 Optimization -- 4 Experiments -- 4.1 Comparisons with Previous Methods -- 4.2 User Study -- 4.3 Ablation Study and Discussion -- 4.4 Application -- 5 Conclusion -- References -- Simple Baselines for Image Restoration -- 1 Introduction -- 2 Related Works -- 2.1 Image Restoration -- 2.2 Gated Linear Units -- 3 Build a Simple Baseline -- 3.1 Architecture -- 3.2 A Plain Block -- 3.3 Normalization -- 3.4 Activation -- 3.5 Attention -- 3.6 Summary -- 4 Nonlinear Activation Free Network -- 5 Experiments -- 5.1 Ablations -- 5.2 Applications -- 6 Conclusions -- References -- Spike Transformer: Monocular Depth Estimation for Spiking Camera -- 1 Introduction -- 2 Related Works -- 2.1 Bio-Inspired Spiking Camera -- 2.2 Image-Based and Event-Based Monocular Depth Estimation -- 2.3 Transformer for Dense Prediction -- 3 Preliminary: Spike Generation Mechanism -- 4 Spike Transformer for Monocular Depth Estimation -- 4.1 Spike Embedding -- 4.2 Spatio-Temporal Transformer Encoder -- 4.3 Decoder for Depth Prediction -- 4.4 Loss Function -- 5 Experiments -- 5.1 Dataset -- 5.2 Implementation Details -- 5.3 Experiment Results -- 5.4 Ablation Studies -- 6 Conclusions -- References -- Improving Image Restoration by Revisiting Global Information Aggregation -- 1 Introduction -- 2 Related Work -- 3 Analysis and Approach -- 3.1 Revisit Global Operations in Image Restoration Tasks -- 3.2 Test-time Local Converter -- 3.3 Extending TLC to Existing Modules -- 3.4 Discussion -- 4 Experiments -- 4.1 Main Results -- 4.2 Size of Local Window -- 4.3 Extensibility and Complexity -- 5 Conclusion -- References. ; Data Association Between Event Streams and Intensity Frames Under Diverse Baselines -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Problem Formulation and Overall Framework -- 3.2 LSparse-Net -- 3.3 SDense-Net -- 3.4 Dense Prediction Layer -- 3.5 Implementation Details -- 4 Applications -- 4.1 Pose Estimation -- 4.2 Depth Estimation -- 4.3 Ablation Study -- 5 Conclusions -- References -- D2HNet: Joint Denoising and Deblurring with Hierarchical Network for Robust Night Image Restoration -- 1 Introduction -- 2 Related Work -- 3 Data Acquisition -- 4 Methodology -- 4.1 Problem Formulation -- 4.2 D2HNet Architecture and Optimization -- 4.3 Data Processing -- 5 Experiment -- 5.1 Implementation Details -- 5.2 Long-short Fusion Method Experiments -- 5.3 Single-Image Denoising and Deblurring Method Experiments -- 5.4 Ablation Study -- 6 Conclusion -- References -- Learning Graph Neural Networks for Image Style Transfer -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Network Overview -- 3.2 Stylization Graph Construction -- 3.3 Deformable Graph Convolution -- 3.4 Loss Function and Training Strategy -- 4 Experiments -- 4.1 Experimental Settings -- 4.2 Results -- 4.3 Ablation Studies -- 4.4 Diversified Stylization Control -- 5 Conclusions -- References -- DeepPS2: Revisiting Photometric Stereo Using Two Differently Illuminated Images -- 1 Introduction -- 2 Related Work -- 3 Understanding PS2: Photometric Stereo Using Two Images -- 3.1 Shape from Shading (SfS) -- 3.2 Photometric Stereo (PS) -- 3.3 The PS2 Problem -- 4 Method -- 4.1 Network Architecture -- 4.2 More on Lighting Estimation: The Light Space Sampling -- 4.3 Network Training -- 4.4 Loss Functions -- 5 Experimental Results -- 5.1 Ablation Studies -- 6 Conclusion -- References -- Instance Contour Adjustment via Structure-Driven CNN -- 1 Introduction -- 2 Related Works. ; 3 Structural Cues -- 4 Structure-Driven CNN -- 4.1 Structure-Driven Convolution -- 4.2 Structure-Driven Attention -- 5 Experiments -- 5.1 Ablation Study -- 5.2 Comparison with Baselines -- 5.3 User Study -- 6 Conclusion -- References -- Synthesizing Light Field Video from Monocular Video -- 1 Introduction -- 2 Related Work -- 3 Monocular LF Video Estimation -- 3.1 Light Field Frame Prediction -- 3.2 Adaptive Tensor-Display Model -- 3.3 Loss Functions -- 3.4 Disocclusion Handling -- 3.5 Supervised Residual Refinement Block -- 3.6 Overall Loss -- 3.7 Implementation Details -- 4 Experiments -- 4.1 Light Field Video Reconstruction -- 4.2 Ablation Study -- 4.3 Variable Baseline LF Prediction -- 5 Discussion -- 6 Conclusion -- References -- Human-Centric Image Cropping with Partition-Aware and Content-Preserving Features -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Overview -- 3.2 Partition-Aware Feature -- 3.3 Content-Preserving Feature -- 3.4 Network Optimization -- 4 Experiments -- 4.1 Datasets and Evaluation Metrics -- 4.2 Implementation Details -- 4.3 Ablation Study -- 4.4 Comparison with the State-of-the-Arts -- 4.5 Analysis of the Partition-Aware Feature -- 4.6 Analysis of the Heatmap -- 5 User Study -- 6 Limitations -- 7 Conclusion -- References -- DeMFI: Deep Joint Deblurring and Multi-frame Interpolation with Flow-Guided Attentive Correlation and Recursive Boosting -- 1 Introduction -- 2 Related Works -- 3 Proposed Method: DeMFI-Net -- 3.1 DeMFI-Netbs -- 3.2 DeMFI-Netrb -- 4 Experiments -- 4.1 Comparison to Previous SOTA Methods -- 4.2 Ablation Studies -- 5 Conclusion -- References -- Neural Image Representations for Multi-image Fusion and Layer Separation -- 1 Introduction and Related Work -- 2 Method Overview -- 2.1 NIRs for Multi-image Fusion -- 2.2 Two-Stream NIRs for Layer Separation -- 3 Applications -- 3.1 Moiré Removal. ; 3.2 Obstruction Removal -- 3.3 Rain Removal -- 3.4 Discussion -- 4 Conclusion -- References -- Bringing Rolling Shutter Images Alive with Dual Reversed Distortion -- 1 Introduction -- 2 Related Works -- 2.1 Video Frame Interpolation -- 2.2 Rolling Shutter Correction -- 3 Joint RS Correction and Interpolation -- 3.1 Problem Formulation -- 3.2 Evaluation Datasets -- 4 Methodology -- 4.1 Pipeline of IFED -- 4.2 Implementation Details -- 5 Experimental Results -- 5.1 Results on Synthetic Dataset -- 5.2 Results on Real-World Data -- 5.3 Ablation Study -- 6 Conclusions -- References -- FILM: Frame Interpolation for Large Motion -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Loss Functions -- 3.2 Large Motion Datasets -- 4 Implementation Details -- 5 Results -- 5.1 Quantitative Comparisons -- 5.2 Qualitative Comparisons -- 5.3 Ablations -- 5.4 Performance and Memory -- 5.5 Limitations -- 6 Conclusions -- References -- Video Interpolation by Event-Driven Anisotropic Adjustment of Optical Flow -- 1 Introduction -- 2 Related Works -- 3 Method -- 3.1 Revisiting Bi-directional Optical Flow VFI Model -- 3.2 Event-Driven Optical Flow Mask -- 3.3 Pipeline for Our Event-Driven Video Interpolation Model -- 4 Experiments -- 4.1 Implementation Details -- 4.2 Datasets -- 4.3 Comparisons with Previous Methods -- 4.4 Ablation Studies -- 5 Conclusion -- References -- EvAC3D: From Event-Based Apparent Contours to 3D Models via Continuous Visual Hulls -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Apparent Contour Event (ACE) -- 3.2 Learning Apparent Contour Events -- 3.3 Event Based Visual Hull -- 3.4 Global Mesh Optimization -- 4 Experiments -- 4.1 Evaluation Metrics -- 4.2 Evaluating Carving Algorithm -- 4.3 Reconstructing Real Objects -- 4.4 Real Objects with Handheld Camera Trajectory -- 5 Conclusions -- References. ; DCCF: Deep Comprehensible Color Filter Learning Framework for High-Resolution Image Harmonization -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Framework Overview -- 3.2 Comprehensible Neural Color Filter Module -- 3.3 High Resolution Assembly Module -- 3.4 Training Loss -- 4 Experiments -- 4.1 Experimental Setups -- 4.2 Implementation Details -- 4.3 Comparison with Baselines -- 4.4 Qualitative Results -- 4.5 Ablation Studies -- 4.6 Comprehensible Interaction with Deep Model -- 5 Conclusion -- References -- SelectionConv: Convolutional Neural Networks for Non-rectilinear Image Data -- 1 Introduction -- 2 Related Work -- 2.1 Graph Convolution Networks -- 2.2 Transfer Learning -- 2.3 Spherical Images, Superpixels, and Texture Maps -- 3 Selection-Based Convolution -- 4 Selection-Based Convolution for Images -- 4.1 Setting Up Image Graphs -- 4.2 Weight Transfer from 2D Convolutions -- 4.3 Handling Larger Kernels -- 4.4 Pooling Operators and Upsampling -- 4.5 Strides, Dilation, Padding -- 4.6 Verification -- 5 Example Non-rectilinear Configurations -- 5.1 Panoramic and Spherical Images -- 5.2 Superpixel Images -- 5.3 Masked Images and Texture Maps -- 6 Results -- 6.1 Spherical Style Transfer -- 6.2 Spherical Segmentation -- 6.3 Superpixel Depth Prediction -- 6.4 Masked Image and 3D Mesh Style Transfer -- 7 Conclusion -- References -- Spatial-Separated Curve Rendering Network for Efficient and High-Resolution Image Harmonization -- 1 Introduction -- 2 Related Works -- 3 Method -- 3.1 Overall Network Structure -- 3.2 Curve Rendering Modules and Its Variants -- 3.3 Loss Function -- 4 Experiments -- 4.1 Implementation Details -- 4.2 Comparison with Existing Methods -- 4.3 Ablation Studies -- 5 Discussion and Real-World Application -- 6 Conclusion -- References -- BigColor: Colorization Using a Generative Color Prior for Natural Images. ; 1 Introduction.
• Titolo autorizzato: Computer Vision – ECCV 2022
• ISBN: 3-031-20071-3
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910629289403321
• Lo trovi qui: Univ. Federico II
• Serie: Lecture notes in computer science.