ZICQ
中 Log in / Sign up
Newsroom Research & Papers #procedural video understanding #online mistake detection #instruction-conditioned learning #egocentric video #multimodal assistants

Joint Modeling Improves Online Mistake and Slip Detection in Procedural Videos

Avatar of Mr.Xu

By Mr.Xu Compiled & Reviewed by Editorial

Published:

中文阅读 (Chinese) English Version
Co-reading Select any sentence in the article to highlight, annotate, or copy a quote card. Join discussion ↓

Summary:The paper reframes procedural assistance as a causal, online decision problem. Using only observed frames, the task instruction, and prior predictions, the system identifies an attempted step, localizes it, and classifies its execution as correct, modified, slip, correction, or addition; omitted steps are resolved only after the video ends. The authors compare a modular pipeline with a joint predictor and repartition EgoPER so that erroneous videos are represented across all splits. Step-segmen


## Task formulation: From action recognition to deciding whether an action is correct The study addresses real-time assistance for egocentric procedural videos. Instead of reviewing an entire recording after the fact, the system must compare incoming observations with the task instructions as the video unfolds. It determines which step the person is attempting, when that attempt occurs, and whether its execution is correct. At time `t`, the model can access only the observed video prefix `V_1:t`, the instruction `I`, and previous predictions `y_ **Source**: [Hugging Face Daily Papers](https://huggingface.co/papers/2611.25567) (2026-11-19) — END —

Tags: #procedural video understanding #online mistake detection #instruction-conditioned learning #egocentric video #multimodal assistants

Editorial & Fact-Checking Note: This article is compiled from primary research, official release documentation, and source papers by the ZICQ Newsroom pipeline with automated entity verification and human editorial review. If you notice any technical inaccuracy, please submit a correction via our corrections policy or email our editorial desk directly.

What's your take on this story?
🚀 Explore & Submit Works

Community Comments

Loading live comments and annotations…