Learning from Online Videos at Inference Time for Computer-Use Agents

• URL: http://arxiv.org/abs/2511.04137v1
• Date: Thu, 06 Nov 2025 07:29:02 GMT
• Title: Learning from Online Videos at Inference Time for Computer-Use Agents
• Authors: Yujian Liu, Ze Wang, Hao Chen, Ximeng Sun, Xiaodong Yu, Jialian Wu, Jiang Liu, Emad Barsoum, Zicheng Liu, Shiyu Chang,
• Abstract summary: We study how to enable computer-use agents to learn from online videos at inference time effectively.<n>We propose a framework that retrieves and filters tutorial videos, converts them into structured demonstration trajectories, and dynamically selects trajectories as in-context guidance.<n>Our framework consistently outperforms strong base agents and variants that use only textual tutorials or transcripts.
• Score: 41.90425060535666
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Computer-use agents can operate computers and automate laborious tasks, but despite recent rapid progress, they still lag behind human users, especially when tasks require domain-specific procedural knowledge about particular applications, platforms, and multi-step workflows. Humans can bridge this gap by watching video tutorials: we search, skim, and selectively imitate short segments that match our current subgoal. In this paper, we study how to enable computer-use agents to learn from online videos at inference time effectively. We propose a framework that retrieves and filters tutorial videos, converts them into structured demonstration trajectories, and dynamically selects trajectories as in-context guidance during execution. Particularly, using a VLM, we infer UI actions, segment videos into short subsequences of actions, and assign each subsequence a textual objective. At inference time, a two-stage selection mechanism dynamically chooses a single trajectory to add in context at each step, focusing the agent on the most helpful local guidance for its next decision. Experiments on two widely used benchmarks show that our framework consistently outperforms strong base agents and variants that use only textual tutorials or transcripts. Analyses highlight the importance of trajectory segmentation and selection, action filtering, and visual information, suggesting that abundant online videos can be systematically distilled into actionable guidance that improves computer-use agents at inference time. Our code is available at https://github.com/UCSB-NLP-Chang/video_demo.

Related papers

• Dual Learning with Dynamic Knowledge Distillation and Soft Alignment for Partially Relevant Video Retrieval [53.54695034420311]
  In practice, videos are typically untrimmed in long durations with much more complicated background content.<n>We propose a novel framework that distills generalization knowledge from a powerful large-scale vision-language pre-trained model.<n>Experiment results demonstrate that our proposed model achieves state-of-the-art performance on TVR, ActivityNet, and Charades-STA datasets.
  arXiv  Detail & Related papers  (2025-10-14T08:38:20Z)
• Watch and Learn: Learning to Use Computers from Online Videos [50.10702690339142]
  Watch & Learn (W&L) is a framework that converts human demonstration videos readily available on the Internet into executable UI trajectories at scale.<n>We develop an inverse dynamics labeling pipeline with task-aware video retrieval, generate over 53k high-quality trajectories from raw web videos.<n>These results highlight web-scale human demonstration videos as a practical and scalable foundation for advancing CUAs towards real-world deployment.
  arXiv  Detail & Related papers  (2025-10-06T10:29:00Z)
• AgentTrek: Agent Trajectory Synthesis via Guiding Replay with Web Tutorials [53.376263056033046]
  Existing approaches rely on expensive human annotation, making them unsustainable at scale.<n>We propose AgentTrek, a scalable data synthesis pipeline that generates web agent trajectories by leveraging publicly available tutorials.<n>Our fully automated approach significantly reduces data collection costs, achieving a cost of just $0.55 per high-quality trajectory without human annotators.
  arXiv  Detail & Related papers  (2024-12-12T18:59:27Z)
• Agent-based Video Trimming [17.519404251018308]
  We introduce a novel task called Video Trimming (VT)<n>VT focuses on detecting wasted footage, selecting valuable segments, and composing them into a final video with a coherent story.<n>AVT received more favorable evaluations in user studies and demonstrated superior mAP and precision on the YouTube Highlights, TVSum, and our own dataset for the highlight detection task.
  arXiv  Detail & Related papers  (2024-12-12T17:59:28Z)
• CML-MOTS: Collaborative Multi-task Learning for Multi-Object Tracking and Segmentation [31.167405688707575]
  We propose a framework for instance-level visual analysis on video frames. It can simultaneously conduct object detection, instance segmentation, and multi-object tracking. We evaluate the proposed method extensively on KITTI MOTS and MOTS Challenge datasets.
  arXiv  Detail & Related papers  (2023-11-02T04:32:24Z)
• Video alignment using unsupervised learning of local and global features [0.0]
  We introduce an unsupervised method for alignment that uses global and local features of the frames. In particular, we introduce effective features for each video frame by means of three machine vision tools: person detection, pose estimation, and VGG network. The main advantage of our approach is that no training is required, which makes it applicable for any new type of action without any need to collect training samples for it.
  arXiv  Detail & Related papers  (2023-04-13T22:20:54Z)
• Learning and Verification of Task Structure in Instructional Videos [85.511888642497]
  We introduce a new pre-trained video model, VideoTaskformer, focused on representing the semantics and structure of instructional videos. Compared to prior work which learns step representations locally, our approach involves learning them globally. We introduce two new benchmarks for detecting mistakes in instructional videos, to verify if there is an anomalous step and if steps are executed in the right order.
  arXiv  Detail & Related papers  (2023-03-23T17:59:54Z)
• Enabling Weakly-Supervised Temporal Action Localization from On-Device Learning of the Video Stream [5.215681853828831]
  We propose an efficient video learning approach to learn from a long, untrimmed streaming video. To the best of our knowledge, we are the first attempt to directly learn from the on-device, long video stream.
  arXiv  Detail & Related papers  (2022-08-25T13:41:03Z)
• Boundary-aware Self-supervised Learning for Video Scene Segmentation [20.713635723315527]
  Video scene segmentation is a task of temporally localizing scene boundaries in a video. We introduce three novel boundary-aware pretext tasks: Shot-Scene Matching, Contextual Group Matching and Pseudo-boundary Prediction. We achieve the new state-of-the-art on the MovieNet-SSeg benchmark.
  arXiv  Detail & Related papers  (2022-01-14T02:14:07Z)
• Transcript to Video: Efficient Clip Sequencing from Texts [65.87890762420922]
  We present Transcript-to-Video -- a weakly-supervised framework that uses texts as input to automatically create video sequences from an extensive collection of shots. Specifically, we propose a Content Retrieval Module and a Temporal Coherent Module to learn visual-language representations and model shot sequencing styles. For fast inference, we introduce an efficient search strategy for real-time video clip sequencing.
  arXiv  Detail & Related papers  (2021-07-25T17:24:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.