Portrait of Artur Kuramshin

Artur Kuramshin

Master's Research - Université de Montréal
Supervisor
Research Topics
Deep Learning
Robotics
Robustness
Vision and Language

Publications

Finetuning VLAs for Dexterous Manipulation: The Right Checkpoint Is All You Need
Vision-Language-Action models have demonstrated strong manipulation capabilities. A limitation of VLA models is that most manipulation datas… (see more)ets focus on parallel-jaw gripper systems, making transfer to multi-finger hands challenging, particularly as teleoperation datasets are harder to collect with multi-fingered robots. In this work, we study how to best adapt VLAs to dexterous hands by pairing supervised fine-tuning (SFT) with diffusion-steering reinforcement learning (DSRL). Using
RoboArena: Distributed Real-World Evaluation of Generalist Robot Policies
Pranav Atreya
Karl Pertsch
Tony Lee
Moo Jin Kim
Arhan Jain
Cyrus Neary
Edward S. Hu
Kanav Arora
Luca Macesanu
Matthew Leonard
Meedeum Cho
Shivin Dass
Tony Wang
Xingfang Yuan
Abhishek Gupta
Dinesh Jayaraman
Kostas Daniilidis
Roberto Martín-Martín
Youngwoon Lee
Percy Liang
Chelsea Finn
Sergey Levine
Task Robustness via Re-Labelling Vision-Action Robot Data