Enregistré dans:
Détails bibliographiques
Auteurs principaux: Tang, Zitian, Krishnan, Rohan Myer, Yu, Zhiqiu, Sun, Chen
Format: Preprint
Publié: 2023
Sujets:
Accès en ligne:https://arxiv.org/abs/2311.18773
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917106703925248
author Tang, Zitian
Krishnan, Rohan Myer
Yu, Zhiqiu
Sun, Chen
author_facet Tang, Zitian
Krishnan, Rohan Myer
Yu, Zhiqiu
Sun, Chen
contents Learning from (procedural) videos has increasingly served as a pathway for embodied agents to acquire skills from human demonstrations. To do this, video understanding models must be able to obtain structured understandings, such as the temporal segmentation of a demonstration into sequences of actions and skills, and to generalize the understandings to novel environments, tasks, and problem domains. In pursuit of this goal, we introduce Spacewalk-18, a benchmark containing two tasks: (1) step recognition and (2) video question answering, over a dataset of temporally segmented and labeled tasks in International Space Station spacewalk recordings. In tandem, the two tasks quantify a model's ability to: (1) generalize to novel domains; (2) utilize long temporal context and multimodal (e.g. visual and speech) information. Our extensive experimental analysis highlights the challenges of Spacewalk-18, but also suggests best practices for domain generalization and long-form understanding. Notably, we discover a promising adaptation via summarization technique that leads to significant performance improvement without model fine-tuning. The Spacewalk-18 benchmark is released at https://brown-palm.github.io/Spacewalk-18/.
format Preprint
id arxiv_https___arxiv_org_abs_2311_18773
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Spacewalk-18: A Benchmark for Multimodal and Long-form Procedural Video Understanding in Novel Domains
Tang, Zitian
Krishnan, Rohan Myer
Yu, Zhiqiu
Sun, Chen
Computer Vision and Pattern Recognition
Learning from (procedural) videos has increasingly served as a pathway for embodied agents to acquire skills from human demonstrations. To do this, video understanding models must be able to obtain structured understandings, such as the temporal segmentation of a demonstration into sequences of actions and skills, and to generalize the understandings to novel environments, tasks, and problem domains. In pursuit of this goal, we introduce Spacewalk-18, a benchmark containing two tasks: (1) step recognition and (2) video question answering, over a dataset of temporally segmented and labeled tasks in International Space Station spacewalk recordings. In tandem, the two tasks quantify a model's ability to: (1) generalize to novel domains; (2) utilize long temporal context and multimodal (e.g. visual and speech) information. Our extensive experimental analysis highlights the challenges of Spacewalk-18, but also suggests best practices for domain generalization and long-form understanding. Notably, we discover a promising adaptation via summarization technique that leads to significant performance improvement without model fine-tuning. The Spacewalk-18 benchmark is released at https://brown-palm.github.io/Spacewalk-18/.
title Spacewalk-18: A Benchmark for Multimodal and Long-form Procedural Video Understanding in Novel Domains
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.18773