Reproduce “SDPose: Exploiting Diffusion Priors for Out-of-Domain and Robust Pose Estimation” 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
Shuang Liang, Jing He, Lejun Liao, Chuanmeizhi Wang, Guo Zhang, Yingcong Chen, Yuan Yuan
SDPose improves cross-domain pose estimation by repurposing a Stable Diffusion U-Net as a feature extractor. It fuses features from multiple U-Net levels and trains a heatmap pose head alongside an image-latent reconstruction branch, using COCO annotations. On unseen HumanArt, its COCO-trained model reaches 71.8 AP versus 67.4 for ViTPose-H at matched resolution; zero-shot means no HumanArt fine-tuning.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.