AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

UniTac: A Unified Multimodal Model for Cross-Sensor Tactile Understanding and Generation

Jiahang Tu, Fengyu Yang, Chenyang Ma, Xihang Yu, Ziyao Zeng, Shaokai Wu, Hanbin Zhao, Zhi Tao, Chao Zhang, Hui Qian, Alex Wong

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add sensor identification objective (ablation)UniTac · PHYSICLEAR-TestAverage accuracy: 57.38 -> 60.61Improved
add dual-level comprehension (ablation)UniTac · PHYSICLEAR-TestAverage accuracy: 26.52 -> 60.61Improved
add dual-level comprehension (ablation)UniTac · Tactile generationSSIM: 0.758 -> 0.836Improved
add DiT projector (ablation)UniTac · Tactile generationSSIM: 0.794 -> 0.836Improved
add sensor-prior sampling instead of vanilla CFG (ablation)UniTac · Tactile generationSSIM: 0.817 -> 0.836Improved
add sensor-prior sampling instead of vanilla CFG (ablation)UniTac · Tactile generationPSNR: 19.49 -> 19.93Improved
add sensor identification objective (ablation)UniTac · Tactile generationPSNR: 19.91 -> 19.93No clear effect
set lambda_sen to 0.1 instead of 0.01 (ablation)UniTac · PHYSICLEAR-TestAverage accuracy: 58.27 -> 60.61Improved
use CLIP semantic encoder instead of VAE pixel encoder (ablation)UniTac · PHYSICLEAR-TestAverage accuracy: 60.55 -> 60.61No clear effect
increase MLLM queries from 16 to 64 (ablation)UniTac · Tactile generationPSNR: 19.08 -> 19.93Improved
increase MLLM queries from 64 to 256 (ablation)UniTac · Tactile generationPSNR: 19.93 -> 19.96No clear effect
add real tactile input to RGB VLA policypi0.5 VLA · Tactile towel selection and graspingOverall success: 20% -> 95%Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy