12.11.2026 IfI Colloquium: Artificial Touch: the Next Frontier for AI

Speaker: 

Prof. Dr. Aviv Tamar
Technion ECE, Israel Institute of Technology, Israel

Date: Thursday, 12 November 2026, 17:15

Location: room BIM 4-033 at the Department of Informatics (IfI), Binzmühlestrasse 11, 8050 Zürich 

Details about the format of the talk shall be checked always just ahead of a certain presentation date: (information here) 

Abstract

AI can currently understand sensory modalities such as vision and audio at a human or super-human level. However, the algorithms for super-human artificial touch remain a mystery. In this talk I will describe our effort to understand artificial touch in the context of artificial medical palpation - an important and challenging problem. For example, in breast cancer, many cases are still detected by palpation, even though mammography and MRI screening are standard. Our approach to understanding palpation is rooted in learning information states: a method for decision making in partially observed domains, with tight connections to learning world models. To collect large scale data for training our models, we manufactured soft objects with a modular design that allows for high diversity. We then palpated them automatically with a robot, yielding the largest tactile dataset to date (over 800 hours). To learn from this data, we developed LESS - a local neural scene representation that is tailored for tactile data. Building on these, we demonstrate the first AI-based tactile imaging of soft objects. I will discuss potential applications of this technology in medicine, prosthetics, and robotic manipulation.

Bio

Aviv Tamar is an associate professor at the Electrical and Computer Engineering department at Technion. His work focuses on reinforcement learning and robot learning. Aviv is the recipient of the Krill prize, an ERC starting grant, and best paper awards at NeurIPS and NSDI.