XinZhi Embodied and Fudan University have released a series of technical reports introducing the N0 series models, which aim to enhance embodied intelligence by incorporating tactile data. The project includes a large dataset of over 30,000 hours of visual-tactile interaction data, along with models like N0-Foundation for unified tactile representation and N0-VTLA for predictive tactile sensing. These advancements are intended to improve robot manipulation capabilities, enabling them to better understand and interact with the physical world through touch, moving beyond purely visual perception. AI
IMPACT Enhances robot manipulation by integrating tactile feedback, potentially improving performance in complex physical tasks and advancing the field of embodied AI.
RANK_REASON The cluster details technical reports and models released by research institutions, focusing on advancements in tactile data for embodied intelligence.
- Franka
- Fudan University
- N0-Foundation
- N0 series
- N0-TWAM
- N0-VTLA
- NeoData
- NeoteAI
- Piper
- UR5e
- Haptic sensing system with active perception
- N0 series models
- NeoData dataset
- XinZhi Embodied
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
