Abhishek Kumar Sinha

I am currently working as a Scientist in Signal and Image Processing Group at Space Applications Center (SAC), Ahmedabad. My research interest broadly spans the application of Artificial Intelligence to solve computer vision problems, including Image and Video Compression and Spatio-spectral Super-Resolution. I am also interested in studying and utilizing the theoretical aspects of Machine Learning to assess the fair behaviour of neural networks. Other than computer vision, I am also interested in general science, especially Physics. Out of my curiosity, I also explore the connection of neural nets to physical principles.

Prior to joining SAC, I graduated with Bachelors in Electronics and Communication Engineering from Indian Institute of Space Science and Technology, where I was advised by Dr. Deepak Mishra during my undergraduate research in the field of Deep Video Compression. I also spent some time working with Dr. Sooraj R. to study the convergence behaviour of particles’ energies in excited quantum wells (Link here).

news

Nov 9, 2024	Paper is accepted in IEEE/CVF Winter Connference on Applications in Computer Vision (WACV) 2025.
Jun 28, 2024	Paper is accepted in International Conference on Pattern Recognition (ICPR) 2024.
Aug 22, 2023	Paper is accepted in British Machine Vision Comference (BMVC) 2023.
Jun 19, 2023	Paper is accepted in ICML SynS & ML Workshop.
Feb 12, 2023	I am serving as a reviewer for ICCV 2023.
Nov 13, 2022	I am serving as a reviewer for CVPR 2023.
Apr 25, 2022	Two papers accepted in CVPR 2022 Workshops, each in NTIRE Workshop and CLIC Workshop.
May 20, 2021	Paper accepted in IEEE National Conference on Communications ( NCC ).

Conferences & Journals reviewed

ECCV 2024, ICML 2023 Syn&ML workshop, ICCV 2023, CVPR (2023, 2024), CVPR Workshops 2022, IEEE Trans. of Signal Processing

Selected Publications

CharDiff: Improving Sampling Convergence via Characteristic Function Consistency in Diffusion Models

Abhishek Kumar Sinha, and S. Manthira Moorthi

In IEEE/CVF Winter Conference on Applications of Computer Vision, 2025

Bib PDF

@inproceedings{chardiff,
  title = {CharDiff: Improving Sampling Convergence via Characteristic Function Consistency in Diffusion Models},
  author = {Sinha, Abhishek Kumar and Manthira Moorthi, S.},
  booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision},
  year = {2025},
}

Enhancing Adversarial Robustness and Combating Uncertainty Bias in Transductive Zero-Shot Learning: A Framework of Pseudo-Bidirectional Alignment

Abhishek Kumar Sinha, Deepak Mishra, and S. Manthira Moorthi

In International Conference on Pattern Recognition (ORAL presentation, Acceptance rate-8.1 %), 2024

Bib PDF

@inproceedings{tzsl_icpr24,
  title = {Enhancing Adversarial Robustness and Combating Uncertainty Bias in Transductive Zero-Shot Learning: A Framework of Pseudo-Bidirectional Alignment},
  author = {Sinha, Abhishek Kumar and Mishra, Deepak and Manthira Moorthi, S.},
  booktitle = {International Conference on Pattern Recognition (ORAL presentation, Acceptance rate-8.1 %)},
  year = {2024},
}

Lips-SpecFormer: Non-Linear Interpolable Transformer for Spectral Reconstruction using Adjacent Channel Coupling

Abhishek Kumar Sinha, and S. Manthira Moorthi

In British Machine Vision Conference (BMVC), 2023

Bib HTML PDF

@inproceedings{lips_specformer_bmvc,
  title = {Lips-SpecFormer: Non-Linear Interpolable Transformer for Spectral Reconstruction using Adjacent Channel Coupling},
  author = {Sinha, Abhishek Kumar and Manthira Moorthi, S.},
  booktitle = {British Machine Vision Conference (BMVC)},
  year = {2023},
}

Improving the Lipschitz stability in Spectral Transformer through Nearest Neighbour Coupling

Abhishek Kumar Sinha

In ICML Workshop on Synergy of Scientific and Machine Learning Modeling, 2023

Bib HTML

@inproceedings{sinha_coupled_attention,
  title = {Improving the Lipschitz stability in Spectral Transformer through Nearest Neighbour Coupling},
  author = {Sinha, Abhishek Kumar},
  booktitle = {ICML Workshop on Synergy of Scientific and Machine Learning Modeling},
  year = {2023},
}

NL-FFC: Non-Local Fast Fourier Convolution for Image Super Resolution

Abhishek Kumar Sinha, S. Manthira Moorthi, and Debajyoti Dhar

In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2022

Bib PDF

@inproceedings{sinha_nlffc,
  title = {NL-FFC: Non-Local Fast Fourier Convolution for Image Super Resolution},
  author = {Sinha, Abhishek Kumar and Manthira Moorthi, S. and Dhar, Debajyoti},
  booktitle = {2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
  pages = {466--475},
  year = {2022},
  publisher = {IEEE,},
}

Self-Supervised Variable Rate Image Compression Using Visual Attention

Abhishek Kumar Sinha, S. Manthira Moorthi, and Debajyoti Dhar

In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2022

Bib PDF

@inproceedings{sinha_comp,
  title = {Self-Supervised Variable Rate Image Compression Using Visual Attention},
  author = {Sinha, Abhishek Kumar and Manthira Moorthi, S. and Dhar, Debajyoti},
  booktitle = {2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
  pages = { 1720--1724},
  year = {2022},
  publisher = {IEEE,},
}