The N-Body Problem: Parallel Execution from Single-Person Egocentric Video
This addresses the challenge of efficient task automation from limited human demonstrations, though it appears incremental as it builds on existing vision-language models and datasets.
The paper tackles the problem of parallelizing complex activities from a single egocentric video by introducing the N-Body Problem, which aims to maximize speed-up while avoiding physical and causal constraints, resulting in a method that boosts action coverage by 45% and reduces various conflict rates by 45-55% for N=2.
Humans can intuitively parallelise complex activities, but can a model learn this from observing a single person? Given one egocentric video, we introduce the N-Body Problem: how N individuals, can hypothetically perform the same set of tasks observed in this video. The goal is to maximise speed-up, but naive assignment of video segments to individuals often violates real-world constraints, leading to physically impossible scenarios like two people using the same object or occupying the same space. To address this, we formalise the N-Body Problem and propose a suite of metrics to evaluate both performance (speed-up, task coverage) and feasibility (spatial collisions, object conflicts and causal constraints). We then introduce a structured prompting strategy that guides a Vision-Language Model (VLM) to reason about the 3D environment, object usage, and temporal dependencies to produce a viable parallel execution. On 100 videos from EPIC-Kitchens and HD-EPIC, our method for N = 2 boosts action coverage by 45% over a baseline prompt for Gemini 2.5 Pro, while simultaneously slashing collision rates, object and causal conflicts by 55%, 45% and 55% respectively.