cs.CVNov 11, 2026

Where to Generate Touch: Do Image Latents Preserve Material Recognition?

Authors: Wanghao Ye, Aarosh Das, Sihan Chen, Yiting Wang, Bowei Tian, Guoheng Sun, Shwai He, Zheyu Shen, 

Organizations: University of Maryland, College Park · Carnegie Mellon University · Texas A&M University · University of Alabama · 

Abstract

Generated touch could extend robot learning beyond recorded contacts. A tactile generator needs a latent space, and we choose one by testing whether a material classifier trained on real GelSight presses still works on encode-decode reconstructions from pretrained image models. On ObjectFolder Real, the classifier loses 6.5 to 9.7 points on reconstructions from three 16x discrete tokenizers and 0.3 points on those from the Stable Diffusion VAE (SD-VAE). The 16x tokenizers again lose far more on Touch and Go and on DIGIT object recognition (YCB-Slide). Retraining the classifier on reconstructions, with no new labels, narrows the discrete gap. SD-VAE reconstructions also keep object identity, with 76.7% object-matching accuracy against 78.9% for real touch and 67.0% for MAGVIT-v2. We therefore select SD-VAE latents and show a photo-to-touch generator trained in them. Code will be released upon acceptance.

Figures & tables

Explore similar work

CardsList
  1. Language-Guided Representation Learning for Robust Cross-Sensor Material Recognition

    Sep 13, 2026Mashood M. Mohsan, Muhayy Ud Din, Binzhao Xu +2Vision-Based Tactile SensingTactile Sensing

  2. VQ-Touch: A Data-Efficient Tactile Generation Framework Across Sensors and Scenarios

    Jul 16, 2026Kailin Lyu, Long Xiao, Jianing Zeng +3Tactile SensingVector-Quantized Variational Autoencoder