Shravan Venkatraman

I am an M.Sc. computer vision student at MBZUAI. I completed my undergrad in computer science at VIT University, Chennai, advised by Dr. Joe Dhanith and Dr. Pandiyaraju.

I am interested in addressing challenges in deep learning and multimodal representation learning, with a focus on large-scale models for robust image and video understanding, generation, and cross-domain reasoning.

Prior to this, I was a research intern at Nagasaki University advised by Dr. Muthu Subash Kavitha. In the summer of 2024, I interned at MedxAI under the mentorship of Dr. Susan Elias and Dr. Sheena Pravin. I also work on industry- and consultancy-funded projects through SPORIC at VIT.

Email  /  CV  /  Google Scholar  /  GitHub

Explicitly modeling camera poses as probability distributions with learnable uncertainties rather than fixed points in SE(3) achieves high-quality reconstruction even with significant pose errors.

    NEED TO ADD BIBTEX
          

Neural encoding of inter-tissue dependencies enables structurally coherent predictions in boundary-dense regions for histopathology segmentation.

              NEED TO ADD BIBTEX
            

Fusing spatial detail with frequency-guided attention cues enables perceptual underwater image enhancement across color-distorted environments.

@misc{FUSION,
  title={FUSION: Frequency-guided Underwater Spatial Image recOnstructioN}, 
  author={Jaskaran Singh Walia and Shravan Venkatraman and Pavithra LK},
  year={2025},
  eprint={2504.01243},
  archivePrefix={arXiv},
  primaryClass={cs.CV},
  url={https://arxiv.org/abs/2504.01243}, 
}
            

Enforcing physical surface properties through PDE constraints yields geometrically accurate neural scene representations from sparse views.

Structuring attention through multi-scale graphs enable transformers to reason across visual hierarchies.

BCSE332P - Deep Learning Lab (Fall 2024)

Yep it's another Jon Barron and Ben Mildenhall website.
Last updated July 2025.