Vid2Coach is an AI-powered system designed to transform standard how-to videos into interactive, wearable task assistants specifically for individuals who are blind or have low vision (BLV). By leveraging multimodal understanding, the system extracts high-level instructions and demonstration details from videos—such as specific tool use or visual cues—and supplements them with accessible workarounds. Key Features of Vid2Coach
Retrieval-Augmented Generation (RAG) for Accessibility: Because general tutorials often lack non-visual instructions, Vid2Coach uses RAG to supplement steps with accessible tips and workarounds, such as using high-contrast cutting boards or cut-resistant gloves. vid2coach top
1. Objectives & Scope
Goal: Automatically analyze short videos of athletes (e.g., baseball pitchers, tennis serves, golf swings, sprinting starts) and generate concise, actionable coaching cues about upper-body/core ("top") mechanics.
Biomechanical Overlays: Using pose estimation algorithms, Vid2Coach projects joint angles, center of gravity, and force vectors onto the raw footage. A high jumper who thinks they are arching their back sees a red line indicating a 15-degree deficiency. The software quantifies the qualitative, turning art into science without stripping away the art’s beauty.
close
TV ProgramaIšsaugoti svetainės nuorodą, kaip programėlę.