Data associated with the publication: Imitatecholec: A multimodal dataset for long-horizon imitation learning in robotic cholecystectomy

ImitateCholec is a publicly available, multi-modal dataset of over 18,000 ex vivo porcine cholecystectomy demonstrations. Each clipping and cutting phase is broken into 17 fine-grained tasks and captured synchronously via stereo endoscopic and wrist-mounted cameras (JPEG frames) together with da Vinci Research Kit kinematics (CSV). Both optimal executions and recovery maneuvers are included, supporting long-horizon imitation learning, workflow modeling, error recognition, and surgical tool pose estimation. (2025-07)