Subhajit Chaudhury

Senior Research Scientist, IBM Research
subho dot chaudhury at outlook dot com
[Google Scholar] [CV] [ LinkedIn]

About me

Hi there! I am Subhajit and I am currently working as a Senior Research Scientist at IBM TJ Watson Research Center, New York. My research interests involve improving factuality in generative language models, text-based reinforcement learning and robust computer vision algorithms. I obtained my Ph.D. degree from the University of Tokyo, where my dissertation focused on the topic of robust deep learning methods against adversarial vulnerabilities. I received my master's degree in Electrical Engineering from IIT Bombay and bachelor's degree in Electrical Engineering from Jadavpur University.

Please feel free to email me if you are interested in my work and/or want to collaborate with me.

Work Experience

  • IBM TJ Watson Research Center (Oct 2021 - Present)
    IBM Research - Tokyo (Apr 2017 - Oct 2021)
    Position: Senior Research Scientist
    I am part of the Foundations of AI group. My research interests include (1) Factuality in generative language modeling, (2) Text-based reinforcement learning, and (3) Neuro-symbolic methods for AI transparency. Previously, I worked on robust computer vision applications for real-world applications. My work on robust activity detection was used by Japan Institute of Sports Sciences (JISS) to train the Japanese national team for Tokyo Olympics.

  • NEC Research Laboratories, Japan (Oct 2014 - Mar 2017)
    Position: Researcher
    At NEC, I worked on deep learning based crack detection from videos. Our crack detection system was used for inspections of road cracks in Texas, USA. Additionally, our system was used to find defects on the runaway in the Nanki-Shirahama airport in Wakayama, Japan.


1. Subhajit Chaudhury, Sarathkrishna Swaminathan, Daiki Kimura, Prithviraj Sen, Keerthiram Murugesan, Rosario Uceda-Sosa, Michiaki Tatsubori, Achille Fokoue, Pavan Kapanipathi, Asim Munawar and Alexander Gray, Learning Symbolic Rules over Abstract Meaning Representations for Textual Reinforcement Learning, ACL 2023

2. Maxwell Crouse, Pavan Kapanipathi, Subhajit Chaudhury, Tahira Naseem, Ramon Fernandez Astudillo, Achille Fokoue, and Tim Klinger, Laziness Is a Virtue When It Comes to Compositionality in Neural Semantic Parsing, ACL 2023

3. Keerthiram Murugesan, Sarathkrishna Swaminathan, Soham Dan, Subhajit Chaudhury, Chulaka Gunasekara, Maxwell Crouse, Diwakar Mahajan, Ibrahim Abdelaziz, Achille Fokoue, Pavan Kapanipathi, Salim Roukos, and Alexander Gray, MISMATCH: Fine-grained Evaluation of Machine-generated Text with Mismatch Error Types, ACL 2023 (in Findings)

4. Heshan Devaka Fernando, Han Shen, Miao Liu, Subhajit Chaudhury, Keerthiram Murugesan, and Tianyi Chen, Mitigating Gradient Bias in Multi-objective Learning: A Provably Convergent Approach, ICLR 2023

5. Subhajit Chaudhury, Sarathkrishna Swaminathan, Chulaka Gunasekara, Maxwell Crouse, Srinivas Ravishankar, Daiki Kimura, Keerthiram Murugesan, Ramon Fernandez Astudillo, Tahira Naseem, Pavan Kapanipathi and Alexander Gray, X-FACTOR: A Cross-metric Evaluation of Factual Correctness in Abstractive Summarization, EMNLP, 2022.

6. Keerthiram Murugesan, Subhajit Chaudhury, and Kartik Talamadupula, Eye of the Beholder: Improved Relation Generalization for Text-based Reinforcement Learning Agents, AAAI, 2022.

7. Subhajit Chaudhury, Prithviraj Sen, Masaki Ono, Daiki Kimura, Michiaki Tatsubori, and Asim Munawar, Neuro-symbolic Approaches for Text-based Policy Learning, EMNLP 2021.

8. Daiki Kimura, Masaki Ono, Subhajit Chaudhury, Ryosuke Kohita, Akifumi Wachi, Don Joven Agravante, Michiaki Tatsubori, Asim Munawar and Alexander Gray, Neuro-symbolic Reinforcement Learning with First-Order Logic, EMNLP 2021.

9. Ryosuke Kohita, Akifumi Wachi, Daiki Kimura, Subhajit Chaudhury, Michiaki Tatsubori, and Asim Munawar. Language-based general action template for reinforcement learning agents, ACL-IJCNLP 2021 (In Findings)

10. Subhajit Chaudhury and Toshihiko Yamasaki, Robustness of Adaptive Neural Network Optimization under Training Noise, IEEE Access, 2021.

11. Subhajit Chaudhury, Hiya Roy, Sourav Mishra, Toshihiko Yamasaki, Adversarial Training Time Attack against Discriminative and Generative Convolutional Models, IEEE Access, 2021.

12. Hiya Roy, Subhajit Chaudhury, Toshihiko Yamasaki and Tatsuaki Hashimoto, Toward Better Planetary Surface Exploration by Mars Orbital Imagery Inpainting, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (JSTARS), 2020.

13. Subhajit Chaudhury, Daiki Kimura, Kartik Talamadupula, Michiaki Tatsubori, Asim Munawar, and Ryuki Tachibana, Bootstrapped Q-learning with Context Relevant Observation Pruning to Generalize in Text-based Games, EMNLP 2020.

14. Subhajit Chaudhury and Toshihiko Yamasaki, Investigating Generalization in Neural Networks under Optimally Evolved Training Perturbations, IEEE ICASSP, 2020.

15. Phongtharin Vinayavekhin, Subhajit Chaudhury, Asim Munawar, Don Joven Agravante, Giovanni De Magistris, Daiki Kimura and Ryuki Tachibana, Focusing on What is Relevant: Time-Series Learning and Understanding using Attention, International Conference on Pattern Recognition (ICPR), 2018.

16. Tadanobu Inoue, Subhajit Chaudhury, Giovanni De Magistris and Sakyasingha Dasgupta, Transfer learning from synthetic to real images using variational auto-encoders for robotic applications, IEEE ICIP, 2018.

17. Subhajit Chaudhury, Gaku Nakano, Jun Takada, Akihiko Iketani, Spatial-temporal motion field analysis for crack detection on concrete surfaces, IEEE WACV 2017.

18. Subhajit Chaudhury, Hiya Roy, Can fully convolutional networks perform well for general image restoration problems?, International Conference on Machine Vision Applications (MVA), 2017.

19. Sourav Saha, Pritha Ganguly, Subhajit Chaudhury. Vision based human pose estimation for virtual cloth fitting. Proceedings of the 2014 Indian Conference on Computer Vision Graphics and Image Processing (ICVGIP)

20. Subhajit Chaudhury, Subhasis Chaudhuri, Volume preserving haptic pottery, 2014 IEEE Haptics Symposium (HAPTICS).

Flag Counter