EgoSteer: A Full-Stack System Towards Steerable Dexterous Manipulation from Egocentric Videos

  • 类型:arxiv
  • 标识:2607.09701
  • 链接:https://arxiv.org/abs/2607.09701
  • 主分类:engineering
  • 形态:method
  • 被引:3
  • 被引来源:Semantic Scholar
  • S2被引:3
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:A full-stack system that scales dexterous VLA pre-training from egocentric human videos and enables data-efficient real-robot post-training, and robustly executes free-form instructions across 45 diverse tasks, demonstrating adherence to user intent amid multiple candidate tasks and generalization.
  • OpenAlex ID:W7168288854
  • OpenAlex DOI:10.48550/arxiv.2607.09701
  • DOI:10.48550/arxiv.2607.09701
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2607.09701
  • OpenAlex更新:2026-07-19
  • 待LLM分类:否
  • 标题中文:EgoSteer:基于第一人称视频的可控灵巧操作系统
  • TLDR中文:全栈系统:扩展来自第一人称人类视频的灵巧 VLA 预训练,实现数据高效的真实机器人后训练,并在 45 个多样化任务上稳健执行自由形式指令,展示了在多个候选任务中对用户意图的遵循与泛化能力。
  • 来源文件:
  • /inbox/tom/_candidates/2026-07-14-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • /inbox/tom/_candidates/2026-07-15-agent-rag-longcontext-candidates.json
  • [OpenAlex backfill]