Reproduce “Video2SwimFish: An Automated Pipeline for Reconstructing Controllable Fish Models and Biological Locomotion from Real Fish Videos” 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
Hangong Chen, Linfeng Cheng, Tahsin Zaman Jilan, Ian Fuller, Lee Caesar, Jiaye Wu, Yantian Zha
Video2SwimFish turns synchronized videos of individual fish into controllable simulated fish with reconstructed shapes, internal skeletons, and swimming policies based on observed body curvature. The authors use a curvature-based action space and evaluate trajectory following, imitation, and simulated ROV capture. Results show success and resemblance can diverge; benchmarks cover small subsets, use approximate hydrodynamics, and do not test sim-to-real transfer.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.