Srikar Yellapragada

I am a PhD student in the Computer Science department at Stony Brook University, advised by Dimitris Samaras. My research is focused on Generative models for Computer Vision. Before this, I was a Software Engineer at Bloomberg LP, where our team built an ETL pipeline for the ingestion of third-party data.

In 2020, I obtained my Master's in Computer Science from NYU, where I worked with Kyunghyun Cho's group on Similarity of Neural Networks, and interned at IBM Watson Health.

Previously, I obtained a B.Tech in Electrical Engineering from Indian Institute of Technology, Hyderabad, where I worked with Sumohana Channappayya on Image Processing. I interned at Video Analytics Lab, IISc Bangalore.

In my free time, I enjoy playing video games and reading fantasy novels. I'm a big fan of strategy games, such as Dota 2 and Age of Empires 2.

News

• [Feb 2025] ZoomLDM has been accepted at CVPR 2025.
• [Dec 2024]LLNL internship research on ViT robustness was accepted at ICASSP 2025.
• [July 2024] ∞-Brush was accepted at ECCV 2024.
• [Feb 2024] Our large image generation research was accepted at CVPR 2024.
• [Oct 2023] PathLDM was accepted at WACV 2024.
• [Aug 2023] Our study on conditional generation with denoiser representations was accepted at BMVC 2023.
• [Aug 2022] Started my PhD at Stony Brook University.

Publications / Pre-prints

ZoomLDM: Latent Diffusion Model for multi-scale image generation Srikar Yellapragada*, Alexandros Graikos*, Kostas Triaridis, Prateek Prasanna ,Rajarsi Gupta, Joel Saltz, Dimitris Samaras CVPR, 2025 Paper / Large image viewer Multiscale diffusion model for histopathology and satellite imagery.

Leveraging Registers in Vision Transformers for Robust Adaptation Srikar Yellapragada, Kowshik Thopalli , Vivek Narayanaswamy, Wesam Sakla, Yang Liu, Yamen Mubarka, Dimitris Samaras , Jayaraman J. Thiagarajan ICASSP, 2025 Paper Registers make ViTs robust

Learned Representation-Guided Diffusion Models for Large-Image Generation Alexandros Graikos*, Srikar Yellapragada*, Minh-Quan Le, Saarthak Kapse , Prateek Prasanna , Joel Saltz, Dimitris Samaras CVPR, 2024 Paper / Project Page / Code We condition latent diffusion models with SSL embeddings and generate large images in histopathology and satellite imagery.

∞-Brush: Controllable Large Image Synthesis with Diffusion Models in Infinite Dimensions Minh-Quan Le* , Alexandros Graikos* , Srikar Yellapragada, Rajarsi Gupta , Joel Saltz , Dimitris Samaras ECCV, 2024 Paper / Project Page / We introduce the first conditional diffusion model in functional space.

PathLDM: Text Conditioned Latent Diffusion Model for Histopathology Srikar Yellapragada*, Alexandros Graikos*, Prateek Prasanna, Tahsin Kurc, Joel Saltz, Dimitris Samaras WACV, 2024 Paper / Project Page / Code We build a text-conditioned LDM for histopathology using GPT-summarized text reports and CLIP embeddings.

Conditional Generation from Unconditional Diffusion Models using Denoiser Representations Alexandros Graikos*, Srikar Yellapragada*, Dimitris Samaras BMVC, 2023 Paper / Code / Poster We condition unconditional models in limited data scenarios using an auxiliary network built upon denoiser representations.

Are the Proposed Similarity Metrics Also a Measure of Functional Similarity? Manikanta Srikar Yellapragada Master's Thesis, 2020 Paper Short answer: NO. Existing representation similarity metrics cannot fully capture the output (functional) similarity of a neural network.

Deep Learning Based Detection of Acute Aortic Syndrome in Contrast CT Images Manikanta Srikar Yellapragada, Yiting Xie, Benedikt Graf, David Richmond, Arun Krishnan, Arkadiusz Sitek ISBI, 2020 Paper

Automatic Diagnosis of Pulmonary Embolism Using an Attention-Guided Framework: A Large-Scale Study Luyao Shi, Deepta Rajan, Shafiq Abedin, Manikanta Srikar Yellapragada, David Beymer, Ehsan Dehghan MIDL, 2020 Paper

Optical Character Recognition (OCR) for Telugu: Database, Algorithm and Application Konkimalla Chandra Prakash, Y.M.Srikar, Gayam Trishal, Souraj Mandal, Sumohana S. Channappayya ICIP, 2018 Paper / Code