Reproduce “Human Image Generation: A Comprehensive Survey” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhen Jia, Zhang Zhang, Liang Wang, Tieniu Tan
The survey organizes human-image generation into data-driven, knowledge-guided, and hybrid paradigms, reviewing more than 200 papers. Data-driven methods learn from images; knowledge-guided methods use explicit human-body priors; hybrids combine such priors with neural generation. The authors describe learned methods as capable of high-fidelity images and 3D-prior methods as enabling pose and shape control, but potentially producing artifacts; differing tasks limit comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.