Object Context
Object context research focuses on understanding how the surrounding environment and relationships between objects influence object perception and interpretation, particularly within computer vision and AI. Current efforts concentrate on improving object detection accuracy by addressing annotation errors in datasets and developing models that integrate both pixel-level and object-level information, often leveraging transformer architectures and multimodal approaches incorporating language understanding. This work is crucial for advancing AI capabilities in tasks such as scene understanding, human-robot interaction, and robotic manipulation, ultimately leading to more robust and contextually aware AI systems.
Papers
March 27, 2024
November 2, 2023
May 29, 2023
July 21, 2022