An article co-authored by Project Professor Yukie Nagai has been published in IEEE Robotics and Automation Letters. The study demonstrates that, by learning forward and inverse tasks in a shared latent space, a robot can perform the corresponding inverse task under novel conditions using only forward demonstrations. This work was conducted in collaboration with Prof. Emre Ugur and his colleagues.
- Serdar Bahar, Fatih Dogangun, Matteo Saveriano, Yukie Nagai, and Emre Ugur, “Task Parameter Extrapolation via Learning Inverse Tasks From Forward Demonstrations,” IEEE Robotics and Automation Letters, 11(11):12655-12662, November 2026.
[10.1109/LRA.2026.3732891]