VTAM: Video-Tactile-Action Models for Complex Physical Interaction Beyond VLAs

TL;DR

VTAM introduces a multimodal framework combining video and tactile data to improve physical interaction modeling, especially in contact-rich scenarios, surpassing visual-only models in stability and precision.

Contribution

The paper presents VTAM, a novel multimodal world model integrating tactile perception with video transformers through efficient finetuning and regularization, enhancing contact-rich manipulation capabilities.

Findings

Achieves 90% success rate in contact-rich tasks

Outperforms baseline by 80% in potato chip pick-and-place

Demonstrates the importance of tactile feedback in physical modeling

Abstract

Video-Action Models (VAMs) have emerged as a promising framework for embodied intelligence, learning implicit world dynamics from raw video streams to produce temporally consistent action predictions. Although such models demonstrate strong performance on long-horizon tasks through visual reasoning, they remain limited in contact-rich scenarios where critical interaction states are only partially observable from vision alone. In particular, fine-grained force modulation and contact transitions are not reliably encoded in visual tokens, leading to unstable or imprecise behaviors. To bridge this gap, we introduce the Video-Tactile Action Model (VTAM), a multimodal world modeling framework that incorporates tactile perception as a complementary grounding signal. VTAM augments a pretrained video transformer with tactile streams via a lightweight modality transfer finetuning, enabling efficient cross-modal representation learning without tactile-language paired data or independent tactile pretraining. To stabilize multimodal fusion, we introduce a tactile regularization loss that enforces balanced cross-modal attention, preventing visual latent dominance in the action model. VTAM demonstrates superior performance in contact-rich manipulation, maintaining a robust success rate of 90 percent on average. In challenging scenarios such as potato chip pick-and-place requiring high-fidelity force awareness, VTAM outperforms the pi 0.5 baseline by 80 percent. Our findings demonstrate that integrating tactile feedback is essential for correcting visual estimation errors in world action models, providing a scalable approach to physically grounded embodied foundation models.