Dexterous Tactile World Model
cs.CV, cs.LG, cs.RO
Submitted: 2026-09-28
Updated: 2026-09-28
Comments: Project page: https://adonis-galaxy.github.io/dtwm-project-page/
Project page: https://adonis-galaxy.github.io/dtwm-project-page/1
License: http://creativecommons.org/licenses/by/4.0/
The gist: World models for manipulation are typically trained from video, yet the events that determine how manipulation unfolds, such as making and releasing contact, are difficult to observe visually and are
Terminology
Abstract
World models for manipulation are typically trained from video, yet the events that determine how manipulation unfolds, such as making and releasing contact, are difficult to observe visually and are often easier to sense through touch. We present the Dexterous Tactile World Model (DTWM), a video world model for future-frame prediction of egocentric manipulation from both observed video and tactile signals from a glove worn on each hand. We condition a pretrained video diffusion transformer on each hand's tactile signal through a zero-initialized residual at the corresponding hand location in the video tokens, while a causal mask prevents predicted frames from accessing future information. Compared with a vision-only model matched in architecture, parameters, and training, DTWM reduces the underestimation of hand motion from 23% to 9%, while reducing the perceptual error in the hand region by 7.4% across three training runs per model. The benefit also increases over the prediction horizon, with the improvement in the later predicted chunks being about 4.1x larger than in the first. DTWM also outperforms other visual-tactile world models under the same setting, and training with touch improves future-frame prediction even when no touch is available at inference. Ablations show that the model benefits from both the magnitude and spatial location of force: replacing the tactile signal with binary contact states, either per hand or per location, increases prediction error. The observed course of the force indicates whether the interaction will persist or change.
Sources
- Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets
- Visuo-Tactile World Models
- ViTacWorld: Scaling Visuo-Tactile World Models for Contact-Rich Robot Manipulation
- OpenLongTail: Generative Scaling of Long-Tail Driving Data
- Blind Dexterous Grasping via Real2Sim2Real Tactile Policy Learning
- FeelWorld: Visuo-Tactile World Model for Hierarchical Contact Prediction and Planning
- Cosmos World Foundation Model Platform for Physical AI
- Decoupling Vision, Language, and Action for Efficient Multi-Task Robot Policies
- Artificial Foveated Perception for Mitigating Shortcut Learning in Robotic Foundation Models
- Wan: Open and Advanced Large-Scale Video Generative Models
- IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models
- Coffee: Controllable Diffusion Fine-tuning
- ContactWorld: What Representations Matter for Vision-Tactile Latent World Models in Contact-Rich Manipulation
- TouchAnything: A Dataset and Framework for Bimanual Tactile Estimation from Egocentric Video
- TouchWorld: A Predictive and Reactive Tactile Foundation Model for Dexterous Manipulation
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models