Computer Vision – ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part III

ISBN-13: 9783031200618 / Angielski / Miękka / 2022 / 755 str.

Shai Avidan; Gabriel Brostow; Moustapha Cissé

Opis

The 39-volume set, comprising the LNCS books 13661 until 13699, constitutes the refereed proceedings of the 17th European Conference on Computer Vision, ECCV 2022, held in Tel Aviv, Israel, during October 23–27, 2022. The 1645 papers presented in these proceedings were carefully reviewed and selected from a total of 5804 submissions. The papers deal with topics such as computer vision; machine learning; deep neural networks; reinforcement learning; object recognition; image classification; image processing; object detection; semantic segmentation; human pose estimation; 3d reconstruction; stereo vision; computational photography; neural networks; image coding; image reconstruction; object recognition; motion estimation.

Spis treści

TOCH: Spatio-Temporal Object-to-Hand Correspondence for Motion Refinement.- LaTeRF: Label and Text Driven Object Radiance Fields.- MeshMAE: Masked Autoencoders for 3D Mesh Data Analysis.- Unsupervised Deep Multi-Shape Matching.- Texturify: Generating Textures on 3D Shape Surfaces.- Autoregressive 3D Shape Generation via Canonical Mapping.- PointTree: Transformation-Robust Point Cloud Encoder with Relaxed

K-D Trees.- UNIF: United Neural Implicit Functions for Clothed Human

Reconstruction and Animation.- PRIF: Primary Ray-Based Implicit Function.- Point Cloud Domain Adaptation via Masked Local 3D Structure Prediction.- CLIP-Actor: Text-Driven Recommendation and Stylization for Animating Human Meshes.- PlaneFormers: From Sparse View Planes to 3D Reconstruction.- Learning Implicit Templates for Point-Based Clothed Human Modeling.- Exploring the Devil in Graph Spectral Domain for 3D Point Cloud Attacks.- Structure-Aware Editable Morphable Model for 3D Facial Detail

Animation and Manipulation.- MoFaNeRF: Morphable Facial Neural Radiance Field.- PointInst3D: Segmenting 3D Instances by Points.- Cross-Modal 3D Shape Generation and Manipulation.- Latent Partition Implicit with Surface Codes for 3D Representation.- Implicit Field Supervision for Robust Non-rigid Shape Matching.- Learning Self-Prior for Mesh Denoising Using Dual Graph Convolutional Networks.- diffConv: Analyzing Irregular Point Clouds with an Irregular View.- PD-Flow: A Point Cloud Denoising Framework with Normalizing Flows.- SeedFormer: Patch Seeds Based Point Cloud Completion with Upsample Transformer.- DeepMend: Learning Occupancy Functions to Represent Shape for Repair.- A Repulsive Force Unit for Garment Collision Handling in Neural Networks.- Shape-Pose Disentanglement Using SE(3)-Equivariant Vector Neurons.- 3D Equivariant Graph Implicit Functions.- PatchRD: Detail-Preserving Shape Completion by Learning Patch Retrieval and Deformation.- 3D Shape Sequence of Human Comparison and Classification Using Current and Varifolds.- Conditional-Flow NeRF: Accurate 3D Modelling with Reliable Uncertainty Quantification.- Unsupervised Pose-Aware Part Decomposition for Man-Made Articulated Objects.- MeshUDF: Fast and Differentiable Meshing of Unsigned Distance Field Networks.- SPE-Net: Boosting Point Cloud Analysis via Rotation Robustness Enhancement.- The Shape Part Slot Machine: Contact-Based Reasoning for Generating 3D Shapes from Parts.- Spatiotemporal Self-Attention Modeling with Temporal Patch Shift for Action Recognition.- Proposal-Free Temporal Action Detection via Global Segmentation Mask Learning.- Semi-Supervised Temporal Action Detection with Proposal-Free Masking.- Zero-Shot Temporal Action Detection via Vision-Language Prompting.- CycDA: Unsupervised Cycle Domain Adaptation to Learn from Image

to Video.- S2N: Suppression-Strengthen Network for Event-Based Recognition

under Variant Illuminations.- CMD: Self-Supervised 3D Action Representation Learning with

Cross-Modal Mutual Distillation