Deepak Sridhar

PhD Student, Computer Vision, Generative Modeling

prof_pic.jpg

San Diego, California

I am a third year PhD student at UCSD with a focus on Computer Vision. I am working under Prof. Nuno Vasconcelos on generative modeling (diffusion models) for multimodal vision problems (images/video/3D content generation). Currently, I am working on a 2D/3D diffusion model framework (link) that offers high fidelity, controllability, modularity, (re)-usability (adapting existing foundational models) and applicability (data augmentation) for various downstream vision tasks. I am also working on some fundamental vision problems such as efficient classification and detection in images and videos.

Outside of Research, I love to go on hiking and exploring the nature. I also love to participate in outdoor adventures.

news

Sep 29, 2024 Gave a oral presentation of my PromptSliders paper at the Unlearning and Model Editing Workshop in ECCV’24.
Sep 25, 2024 Paper on “Adapting Diffusion Models for Improved Prompt Compliance and Controllable Image Synthesis” accepted to NeurIPS’24! :sparkles:.
Aug 16, 2024 Paper on “Prompt Sliders for Fine-Grained Control, Editing and Erasing of Concepts in Diffusion Models” accepted to ECCV’24-Unlearning and Model Editing Workshop! :sparkles:.
Dec 1, 2023 New preprint available on arXiv! SCHEME: Scalable Channel Mixer for Vision Transformers.
Sep 26, 2022 Started my PhD in Fall 2022 with Jacobs Fellow Award (Highest recognition in ECE at UCSD).

selected publications

2024

  1. Adapting Diffusion Models for Improved Prompt Compliance and Controllable Image Synthesis
    Deepak Sridhar, Abhishek Peri, Rohit Rachala, and 1 more author
    In Neural Information Processing Systems, 2024
  2. Prompt Sliders for Fine-Grained Control, Editing and Erasing of Concepts in Diffusion Models
    Deepak Sridhar, and Nuno Vasconcelos
    In In Proceedings of the IEEE/CVF European Conference on Computer Vision Workshops, 2024

2023

  1. SCHEME: Scalable Channel Mixer for Vision Transformers
    Deepak Sridhar, Yunsheng Li, and Nuno Vasconcelos
    2023

2021

  1. Class semantics-based attention for action detection
    Deepak Sridhar, Niamul Quader, Srikanth Muralidharan, and 3 more authors
    In In Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021