Computer Vision - ECCV 2024

The multi-volume set of LNCS books with volume numbers 15059 up to 15147 constitutes the refereed proceedings of the 18th European Conference on Computer Vision, ECCV 2024, held in Milan, Italy, during September 29October 4, 2024.

The 2387 papers presented in these proceedings were carefully reviewed and selected from a total of 8585 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; motion estimation.

Inhalt

Towards Latent Masked Image Modeling for Self-Supervised Visual Representation Learning.- Nuvo: Neural UV Mapping for Unruly 3D Representations.- Towards High-Quality 3D Motion Transfer with Realistic Apparel Animation.- AnyHome: Open-Vocabulary Large-Scale Indoor Scene Generation with First-Person View Exploration.- Better Call SAL: Towards Learning to Segment Anything in Lidar.- DGInStyle: Domain-Generalizable Semantic Segmentation with Image Diffusion Models and Stylized Semantic Control.- DECOLLAGE: 3D Detailization by Controllable, Localized, and Learned Geometry Enhancement.- Scene-aware Human Motion Forecasting via Mutual Distance Prediction.- FSGS: Real-Time Few-shot View Synthesis using Gaussian Splatting.- Open Panoramic Segmentation.- iMatching: Imperative Correspondence Learning.- COSMU: Complete 3D human shape from monocular unconstrained images.- MAP-ADAPT: Real-Time Quality-Adaptive Semantic 3D Maps.- Appearance-based Refinement for Object-Centric Motion Segmentation.- SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidance.- Open Vocabulary Multi-Label Video Classification.- Optimal Transport of Diverse Unsupervised Tasks for Robust Learning from Noisy Few-Shot Data.- Regularizing Dynamic Radiance Fields with Kinematic Fields.- MICDrop: Masking Image and Depth Features via Complementary Dropout for Domain-Adaptive Semantic Segmentation.- Efficient Pre-training for Localized Instruction Generation of Procedural Videos.- MTKD: Multi-Teacher Knowledge Distillation for Image Super-Resolution.- DEAL: Disentangle and Localize Concept-level Explanations for VLMs.- Fast Encoding and Decoding for Implicit Video Representation.- Surf-D: Generating High-Quality Surfaces of Arbitrary Topologies Using Diffusion Models.- Diffusion-Refined VQA Annotations for Semi-Supervised Gaze Following.- IMMA: Immunizing text-to-image Models against Malicious Adaptation.- Motion-Oriented Compositional Neural Radiance Fields for Monocular Dynamic Human Modeling.