
200个任务、1700万帧!大晓ACE-Data-0把真实家庭场景变成物理智能「数据引擎」
Quick Answer
Daxiao Robotics and NTU S-Lab launched ACE-Data-0, a multimodal dataset with 15 hours of data, 17 million frames, and 200 tasks, enhancing embodied AI learning in real-world scenarios.
Quick Take
This L5 dataset captures high-fidelity human-object interactions and provides a robust foundation for physical intelligence models, addressing gaps in existing models' performance in complex tasks.
Key Points
- ACE-Data-0 includes 150 hours of data and 17 million video frames.
- Dataset features 200 task categories and 50 participants in real home environments.
- Captures multimodal signals including video, audio, and tactile feedback.
- Establishes a three-tier evaluation benchmark for embodied perception.
- Addresses existing models' limitations in contact perception and long-term task execution.
Source Excerpt
大晓ACE-Data-0把真实家庭场景变成物理智能「数据引擎」
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from 雷峰网 AI
See more →
刚刚,GPT 5.6 发布会上,OpenAI 暴露了哪些 Agent 技术路线?
OpenAI's GPT 5.6 integrates ChatGPT and Codex, introducing a for complex task execution, with models Soul, Terra, and Luna for efficient workflow management. The release emphasizes task orchestration, contextual understanding, and robust security measures for enterprise applications.

