Samyak Rawlekar

I am a PhD student in the Computer Vision and Robotics Laboratory at the University of Illinois Urbana-Champaign. I am advised by Prof. Narendra Ahuja.

My research focuses on visual and vision-language representations for scene understanding and physical interaction. My interests include:

  • Self-supervised representations: Analyzing features from pretrained vision models to understand how they represent objects and their parts.
  • Vision-language learning: Adapting vision-language models for multi-object recognition, segmentation, and multimodal retrieval.
  • 3D perception and physical interaction: Identify the parts of a 3D object that support an action shown in an image or described in text.

Previously, I obtained MS in Computer Engineering from New York University, and B.Tech in Electrical Engineering from IIT Dharwad.

Profile photo

Selected Publications

Please visit my Google Scholar for the full list of publications


PartDINO Teaser

Complementing DINO Features with Image Structure for Part Discovery

ObjectDINO Teaser

Finding Distributed Object-Centric Properties in Self-Supervised Transformers

DCLIP Teaser

Efficiently Disentangling CLIP for Multi-Object Perception

PositiveCoOp Teaser

PositiveCoOp: Rethinking Prompting Strategies for Multi-Label Recognition with Partial Annotations

MLR GCN Teaser

Improving Multi-label Recognition using Class Co-Occurrence Probabilities

S30 Teaser

S3O: A Dual-Phase Approach for Reconstructing Dynamic Shape and Skeleton

LIMR Teaser

LIMR: Learning Implicit Representation for Reconstructing Articulated Objects

Template credit