Summary

This repository serves as a curated, daily-updated tracker for the latest advancements in 3D Gaussian Splatting research, sourced from arXiv. It aims to keep researchers and practitioners informed about cutting-edge papers, projects, and resources in this rapidly evolving field, providing tools for efficient discovery and management of relevant literature.

How It Works

The project automates the collection and categorization of 3D Gaussian Splatting research papers from arXiv on a daily basis, ensuring users stay abreast of the latest developments. It employs intelligent parsing to extract key metadata, classify papers into over 14 predefined research topics (such as Acceleration, Dynamic Scenes, SLAM, Avatar Generation), and automatically identify associated external resources like GitHub repositories, project pages, datasets, and demos. The system is designed for robust operation with multi-layer error handling and CI/CD integration via GitHub Actions, guaranteeing consistent daily updates.

Quick Start & Requirements

• Installation: Clone the repository and install dependencies using pip install -r requirements.txt.
• Setup: Run python main.py init for an interactive wizard to configure search keywords, arXiv domains (e.g., cs.CV, cs.GR), time ranges, and optionally an OpenAI-compatible API key for LLM-powered keyword suggestions.
• Prerequisites: Python environment, requirements.txt dependencies. An OpenAI-compatible API key is recommended for the LLM suggestion feature.

Highlighted Details

• Unified CLI & Interactive Setup: A single main.py entry point simplifies operations with subcommands for initialization, searching, keyword suggestion, BibTeX export, and README generation. An interactive wizard guides users through setting up search parameters and API keys.
• Advanced Search & Filtering: Supports custom keyword configurations (title, abstract, or both), specific arXiv domain filtering (e.g., cs.CV, cs.GR), and flexible time range filtering (relative periods like 6m, 1y, or absolute date ranges).
• Smart Link Extraction & BibTeX Export: Automatically extracts and classifies various resource links from paper abstracts. It also fetches and exports BibTeX entries directly from arXiv, with options for filtering by category and date for easy citation management.
• LLM Keyword Suggestion: Integrates with OpenAI-compatible APIs to generate optimized search keywords from paper titles or arXiv IDs, enhancing literature discovery.
• Automated Paper Collection & Categorization: Daily crawling via GitHub Actions ensures the list is current, with papers intelligently categorized into 14+ topics.

Maintenance & Community

The project is maintained through automated daily updates via GitHub Actions. Contribution guidelines are provided for submitting pull requests to improve the curated list.

Licensing & Compatibility

The README does not specify a software license. Users should verify licensing for any included code or data.

Limitations & Caveats

The accuracy and completeness of the tracked research are dependent on arXiv's data and categorization. The LLM keyword suggestion feature requires access to an OpenAI-compatible API and may incur costs.