Researchers from IIIT Hyderabad have achieved significant recognition at the 2026 CVPR conference, a premier event in AI. Their innovative work in creating ‘lighter’ 3D models and enhancing video comprehension with smaller datasets won them top honors. The institute’s Centre for Visual Information Technology received awards for best paper and best paper runner-up, among other presentations. Darshan Singh’s award-winning research focused on improving video understanding through structured descriptions, while Kunal Bhosikar’s work simplified AI-generated 3D models for efficiency. Additionally, Bhosikar introduced a method to protect photos from unauthorized 3D reconstructions. Other notable contributions included Vaibhav Agrawal’s SeeThrough3D, improving AI’s understanding of occlusions in images, and Prof. Ravi Kiran’s RoadTones, which adapts road incident descriptions based on the audience. The institute also showcased research in scientific communication and document intelligence, with a previous paper receiving the EgoVis distinguished paper award. Professor C.V. Jawahar delivered a keynote address at the event.
Did IIIT Hyderabad achieve top recognition at the CVPR AI Conference?
