Track-and-Complete: Learning Humanoid Skills from a Single Failed Human Video

Learning humanoid skills from videos typically requires a successful human demonstration, which often demands custom data collection. Although failures have traditionally been treated only as negative examples in robot learning, they can still reveal a usable trajectory prefix before the task fails, as well as the intended outcome. To leverage this information from a failed-attempt video, we propose $TRACC$, a pipeline that imitates the useful half of the motion trajectory and then completes the task based on the inferred task outcome. The usable motion prefix serves as prior knowledge until the failure occurs, after which the task-completion reward guides the policy to learn the intended task goal without relying on a successful task trajectory. We evaluate our method on the four in-the-wild failed human tasks from the Oops! dataset. Our experiments demonstrate the effectiveness of the proposed approach for learning from failed attempts when no successful demonstration is available. Thus, we provide a baseline for learning humanoid skills from a failed human video demonstrations. Code will be released.