Hongyu Li

NVIDIA

収録論文 15本 ・ フィジカルAI/ロボット学習

※arXiv著者名で収集。同姓同名の別人の論文が含まれる場合があります。

論文

• Hydra-0: 汎用世界モデリングと制御のためのアクションフロー
  世界モデル2026/8/18

  ロボットの動作をピクセル運動として表現するアクションフローを用いた汎用世界モデルを提案し、動作誤差を大幅に低減し、ゼロショット合成やデータ効率的な適応を実現した。

• Deform360: A Massive Multi-view Visuotactile Dataset for Deformable World Models
  2026/7/1
• NovaPlan: Zero-Shot Long-Horizon Manipulation via Closed-Loop Video Language Planning
  2026/2/1
• From Instruction to Event: Sound-Triggered Mobile Manipulation
  2026/1/1
• NovaFlow: Zero-Shot Manipulation via Actionable Flow from Generated Videos
  2025/10/1
• UniTac: Whole-Robot Touch Sensing Without Tactile Sensors
  2025/7/1
• ViTa-Zero: Zero-shot Visuotactile Object 6D Pose Estimation
  2025/4/1
• V-HOP: Visuo-Haptic 6D Object Pose Tracking
  2025/2/1
• HyperTaxel: Hyper-Resolution for Taxel-Based Tactile Signals Through Contrastive Learning
  2024/8/1
• ODTFormer: Efficient Obstacle Detection and Tracking with Stereo Cameras Based on Transformer
  2024/3/1
• StereoNavNet: Learning to Navigate using Stereo Cameras with Auxiliary Occupancy Voxels
  2024/3/1
• E(2)-Equivariant Graph Planning for Navigation
  2023/9/1
• ViHOPE: Visuotactile In-Hand Object 6D Pose Estimation with Shape Completion
  2023/9/1
• StereoVoxelNet: Real-Time Obstacle Detection Based on Occupancy Voxels from a Stereo Camera Using Deep Neural Networks
  2022/9/1
• Deep Reinforcement Learning based Robot Navigation in Dynamic Environments using Occupancy Values of Motion Primitives
  2022/8/1