This repository is a curated collection of resources, primarily research papers and associated code, focused on conditional content generation. It targets researchers and practitioners in AI, particularly those working on human motion, image, and video synthesis driven by various conditions like text, audio, or music. The benefit is a centralized, up-to-date overview of the state-of-the-art in this rapidly evolving field.

How It Works

The repository organizes papers by generation modality (motion, image, video) and conditioning type (text, audio, music). It leverages diffusion models extensively, highlighting their application in generating diverse and controllable content. The structure allows users to quickly find relevant research and understand the landscape of conditional generation techniques.

Highlighted Details

• Extensive coverage of diffusion models for human motion synthesis, including text-driven, music-driven, and audio-driven approaches.
• Includes papers on conditional image generation, such as text-to-image, sketch-guided, and image editing.
• Features research on text-to-video generation and 3D content generation from text.
• Many listed papers include links to their corresponding code repositories, facilitating reproducibility and further research.

Maintenance & Community

The repository is maintained by Haofan Wang, who welcomes academic collaboration and internship inquiries from individuals with published top-tier conference papers.

Licensing & Compatibility

The repository itself is a collection of links and information; licensing for individual papers or code repositories is not specified here and would need to be checked on a per-project basis.

Limitations & Caveats

This is a curated list of research papers and not a software library; it does not provide direct implementation or runnable code for the described methods. The focus is on academic resources, and practical implementation details or ease of use are not evaluated.