Abeer Banerjee

Research

My research interests lie at the intersection of computational imaging, computer vision, and machine learning. I focus on computational image formation and reconstruction in low-data environments. Currently, I am working on the computational reconstruction of optoacoustic microscopic images. As a DAAD Bi-national PhD Fellow, I am under the joint supervision of Dr. Sanjay Singh and Prof. Vasilis Ntziachristos.

Selected Publications

I have included a selection of my peer-reviewed publications primarily on computational imaging and also on some other related topics.

Towards Physics-informed Cyclic Adversarial Multi-PSF Lensless Imaging

IEEE Transactions on Emerging Topics in Computational Intelligence, 2025

Impact factor: 6.5 - JCR Q1

Abeer Banerjee, Sanjay Singh

We propose a physics-informed cyclic adversarial framework that explicitly accounts for multiple point spread functions for lensless image reconstruction. The method improves robustness under unknown optical misalignment and diffuser perturbations.

[Paper] [arXiv]

Towards Lensless Image Deblurring with Prior-Embedded Implicit Neural Representations

Engineering Applications of Artificial Intelligence, Elsevier, 2025

Impact factor: 8.0 | JCR Q1

Abeer Banerjee, Sanjay Singh

We propose prior-embedded INRs for lensless image deblurring. We show that even with one domain restricted image for embedding priors, we can have superior reconstructions compared to other iterative neural reconstruction methods.

[Paper] [arXiv]

Reconstructing Synthetic Lensless Images in the Low-Data Regime

British Machine Vision Conference (BMVC), 2023

CORE Rank (2026): A | H5-index (2026): 57

Abeer Banerjee, Himanshu Kumar, Sumeet Saurav, Sanjay Singh

We perform domain restricted reconstruction of synthetic lensless images using underparameterized neural networks. We show that reconstructions can be achieved with only 10-15 domain restricted images using a two-step approach, domain-restricted pre-training and PSF-aware iterative reconstruction.

[Paper] [Code]

Physics-informed Deep Deblurring: Over-parameterized vs. Under-parameterized

International Conference on Image Processing (ICIP), 2023 (Oral)

CORE Rank (2026): B | H5-index (2026): 55

Abeer Banerjee, Sumeet Saurav, Sanjay Singh

We propose a physics-informed deep deblurring framework that compares over-parameterized and under-parameterized neural networks for lensless image deblurring. We show that under-parameterized networks can be more robust and efficient in certain regimes.

[Paper] [Code]

Lensless Image Reconstruction with an Untrained Neural Network

Image and Vision Computing New Zealand (IVCNZ), 2022

CORE Rank (2026): C | H5-index (2026): 16

Abeer Banerjee, Himanshu Kumar, Sumeet Saurav, Sanjay Singh

[Paper]

ParaColorizer: Realistic Image Colorization using Parallel Generative Networks

The Visual Computer Journal, Springer, 2023

Impact factor: 3.5 | JCR Q2

Himanshu Kumar*, Abeer Banerjee*, Sumeet Saurav, Sanjay Singh (*Equal Contribution)

[Paper] [arXiv] Copyright Registration No. SW-17058/2023

Generalized Gaze-Vector Estimation in Low-light with Encoded Event-driven Neural Network

International Joint Conference on Neural Networks (IJCNN), 2024 (Oral)

CORE Rank (2026): B | H5-index (2026): 66

Abeer Banerjee, Naval K Mehta, Shyam S Prasad, Himanshu Kumar, Sumeet Saurav, Sanjay Singh

We propose a novel event-driven neural network architecture for low-light gaze estimation. We captured a novel Gaze-FELL dataset and developed a temporal event-encoding scheme for a neural network driven gaze estimation and attained a 75-pixel acccuracy of 97%.

[Paper] [arXiv]

Gaze Detection using Encoded Retinomorphic Events

International Conference on Intelligent Human Computer Interaction (IHCI), 2022 (Oral)

H5-index (2026): 14

Abeer Banerjee, Shyam S Prasad, Naval K Mehta, Himanshu Kumar, Sumeet Saurav, and Sanjay Singh

[Paper]

JS-SpoofNet: A Jointly Supervised Parallel Branched Neural Network for Spoof Detection

Neurocomputing Journal, Elsevier, 2023

Impact factor: 6.0 | JCR Q1

Shyam S Prasad, Naval K Mehta, Abeer Banerjee, Sumeet Saurav, Sanjay Singh

[Paper] Copyright Registration No. SW-16929/2023

A Multimodal Dataset for Enhancing Industrial Task Monitoring and Engagement Prediction

International Conference on Human-Robot Interaction (HRI), 2025

CORE Rank (2026): A* | H5-index (2026): 52

Naval K Mehta, Arvind, Himanshu Kumar, Abeer Banerjee, Sumeet Saurav, Sanjay Singh

[Paper] [arXiv] [Code]

Voluntary Services