Analog Real Robots (ARR)

Valts Blukis, Balakumar Sundaralingam, Alex Zook, Stan Birchfield, Jonathan Tremblay

Analog Real Robots (ARR) is a framework that combines Vision-Language Models (VLMs) and Large Language Models (LLMs) to enable dynamic scene understanding and adaptive task execution for robotic manipulation in complex environments. It excels in recognizing unseen objects, reasoning spatially, and generating robust, adaptable plans, outperforming existing methods.

Paper

Download Paper

Video

Citation

@article{blukis2024analog, author = {Blukis, Valts and Sundaralingam, Balakumar and Zook, Alex and Birchfield, Stan and Tremblay, Jonathan}, title = {Analog Real Robots}, year = {2024}, url = {analogrealrobots.pages.dev}, howpublished = {\url{analogrealrobots.pages.dev}}, note = {Accessed: 2024-11-26} }