Arthur Szlam

Papers

Total Citations

H-Index

About

Arthur Szlam is a leading researcher in machine learning and robotics, with key contributions in representation learning, semantic scene understanding, and multimodal AI. He is best known for pioneering CLIP-Fields, a weakly supervised framework that maps spatial locations to semantic embedding vectors, enabling robots to perform segmentation, instance identification, semantic search, and view localization within 3D environments. This work, published in 2022 and 2023, has garnered nearly 80 citations, reflecting its impact on bridging vision-language models with robotic memory. Szlam’s research addresses fundamental challenges in grounding abstract semantic concepts in physical space, allowing robots to reason about their surroundings without extensive manual annotations. His contributions have advanced the field of embodied AI, making robotic systems more adaptable and context-aware. Beyond CLIP-Fields, Szlam has collaborated on foundational work in deep learning architectures and optimization, influencing both academic research and practical applications. His ability to integrate cutting-edge techniques from natural language processing and computer vision into robotics has positioned him as a key figure in the development of intelligent, autonomous systems.