site.btaINSAIT Participates in Leading AI Robotics Conference in South Korea

A paper by the Institute for Computer Science, Artificial Intelligence and Technology (INSAIT) at Sofia University will be presented at the Conference on Robot Learning 2025, the INSAIT press service reported on Monday. The forum is taking place in Seoul, South Korea, between September 27 and 30.

The paper introduces MotoVLA, a novel approach that enables robots to acquire manipulation skills from unlabelled human and robot videos, drastically reducing the need for massive action-labelled datasets. By predicting 3D dynamics directly from video and aligning them to actions with only a small labelled set, MotoVLA unlocks out-of-action generalization, which would allow robots to perform completely new tasks without ever seeing their action labels during training.

The research team includes INSAIT researchers Alexander Spiridonov, Dr Jan-Nico Zaech, Nikolay Nikolov, Prof. Luc Van Gool, and Dr Danda Paudel.