aifollow.news
Back Rohan Paul
Rohan Paul· @rohanpaul_ai · X· · Original publication time AI score40

EgoHome introduces an open benchmark for first-person household video understandingMachine translation

AI introduction

EgoHome uses human-annotated, first-person household-chore video, with hand boxes, subtasks, hand actions and hand captions on every clip. It includes more than 750 segments shorter than 500 milliseconds; the data, metrics and judge prompts are open.

Original post · AI translation

@rohanpaul_ai
Translation · refer to the original

@perceptroninc 哇,没想到 GBT6 S-Tri 在这类工作上这么擅长。 世界需要密集的第一人称视角标注。

ReplyPerceptron AI@perceptroninc
Translation
隆重推出 EgoHome:一个用于密集、亚秒级第一人称视角理解的全新开放基准。 由人工标注的家务视频,每个片段都有 4 种标注类型:手部边界框、子任务、手部动作、手部文字描述。 其中有超过 750 个时长不到 500 毫秒的片段。 数据、指标和评判提示词全部开放。🧵
Original

Introducing EgoHome: a new open benchmark for dense, sub-second egocentric understanding. Human-annotated household-chore video with 4 annotation types on every clip: hand boxes, subtasks, hand actions, hand captions. It features over 750 segments under 500ms. Data, metrics, and judge prompts are all open. 🧵

View this post on X ↗

Source: Rohan Paul · x.com

Research
Found an error? Send a correction