tlabel (TouchLabel AI)

tlabel (TouchLabel AI) is a Python package that helps robotics researchers load and label tactile sensor data with zero setup. Tactile data from touch sensors is harder to work with than images, and tlabel gives researchers a straightforward way to import, review, and annotate it before feeding it into downstream models or analysis. It supports data from several sensor sources, including GelSight Mini, DIGIT, and PaXini, so labs using different hardware can rely on one tool rather than building custom loaders for each device. Because it is distributed through PyPI, tlabel installs like any other Python dependency and fits into existing research pipelines. Its scope is deliberately focused on the loading-and-labeling stage rather than full model training, and it assumes users are comfortable working in Python with tactile data, making it a fit for robotics research teams rather than general users.

Robotics researchers who need to load and annotate tactile sensor data from hardware like GelSight Mini, DIGIT and PaXini without setup overhead.