This repository serves as a curated collection of papers and their associated open-source code from the ECCV 2024 computer vision conference. It aims to provide researchers and practitioners with a centralized resource for the latest advancements in various computer vision subfields, facilitating reproducibility and further research.

How It Works

The repository organizes ECCV 2024 papers by topic, including areas like 3D Gaussian Splatting, State Space Models (Mamba), Avatars, CLIP, MAE, Embodied AI, GANs, GNNs, Multimodal Large Language Models (MLLM), Large Language Models (LLM), NAS, OCR, NeRF, DETR, Prompting, Diffusion Models, ReID, Long-Tail distributions, Vision Transformers, Vision-Language tasks, Self-supervised learning, Data Augmentation, Object Detection, Anomaly Detection, Visual Tracking, Semantic Segmentation, Instance Segmentation, Panoptic Segmentation, Medical Imaging, Video Object Segmentation, Autonomous Driving, 3D Point Clouds, Image Editing, Low-level Vision, Image Generation, Video Generation, Video Understanding, Action Recognition, Knowledge Distillation, Image Compression, Stereo Matching, Scene Graph Generation, and Counting. Each entry typically links to the paper's abstract, project page, and code repository.

Quick Start & Requirements

To utilize the resources, users should navigate to the specific paper or project of interest within the categorized list. Each linked code repository will detail its own installation and execution instructions, often involving Python environments (e.g., conda, pip), deep learning frameworks (PyTorch, TensorFlow), and potentially specific hardware requirements like GPUs with CUDA.

Highlighted Details

• Comprehensive coverage of ECCV 2024 papers across a wide spectrum of computer vision domains.
• Direct links to official project pages, arXiv papers, and GitHub repositories for each listed work.
• Categorization facilitates easy discovery of state-of-the-art methods in specific research areas.
• Community-driven updates are encouraged via GitHub issues for new submissions.

Maintenance & Community

The repository is maintained by amusi and welcomes community contributions through GitHub issues to add new ECCV 2024 papers and projects. Further resources for past CVPR and ECCV conferences, as well as general computer vision paper discussions, are linked.

Licensing & Compatibility

The licensing of individual projects varies, as each paper's code is hosted on its respective repository. Users must consult the license of each linked project for usage rights and compatibility, particularly for commercial applications.

Limitations & Caveats

This repository is a curated list and does not host the code or papers directly. The availability and maintenance of the linked code repositories are dependent on the original authors. Some entries may lack direct code links or have incomplete information.