Avirup Dey

I am a 3rd-year Computer Science PhD student at the University of Bath, advised by Prof. Vinay Namboodiri. In Summer 2025, I completed a research internship at SpeechGraphics where I worked on real-time audio-driven photorealistic avatars.

I completed my B.E. in Electronics and Communication Engineering in 2023, from Jadavpur University, India. As an undergraduate, I was a research assistant at DVLP-CMATER Lab under Prof. Nibaran Das, where I worked on deep learning-based models for document analysis and cell imaging. As a recipient of the Mitacs Globalink Research Internship, I spent the Summer of 2022 at Ontario Tech University, where I worked under Prof. Mehran Ebrahimi on MRI Super-Resolution. I have also worked as a research assistant at the IVPR Group under Prof. Ananda Shankar Chowdhury, where I worked on domain adaptation methods and event cameras.

Email  /  Twitter  /  Scholar  /  Github

My research spans Computer Vision and Computer Graphics, with a focus on building expressive and controllable digital humans. I am broadly interested in disentangled representations that enable flexible animation, editing, and performance transfer. My recent work includes problems such as audio-driven facial animation, cross-domain retargeting, and blendshape modelling.
Previously, as an undergrad, I worked on inverse imaging problems. Check out my recent publications below.

TLDR: Learnable localized blendshapes that replace PCA-based ones in off-the-shelf models for fine-grained control.

TLDR: Adapting diffusion models for cross-domain expression retargeting without using multi-domain data.

TLDR: Pilot study on cross-modal MRI super-resolution. Optimized off-the-shelf i2i models to leverage sparsity of the scans.

TLDR: GAN-based augmentation helps segmentation models perform better on document binarization benchmarks (DIBCO).

TLDR: Expressing kernel convolution as a matrix-matrix product and reducing image rank boosts training efficiency.

© Avirup Dey. Template borrowed from John Barron