Abhra Chaudhuri

I have now joined the Fujitsu Research of Europe as a Senior Researcher.

My PhD work was supervised by Anjan Dutta and Zeynep Akata. During my PhD, I was an Enrichment Student at The Alan Turing Institute, and an Intern at the Fujitsu Research of Europe.

My area of interest in research is learning theory, presently with a focus at the intersection of causality and geometric deep learning. At Fujitsu, my work involves multi-modal learning on geometric domains (like graphs), with a focus on interpretability.

My PhD research was on representation learning from multiple views, primarily for problems in computer vision, with the central objective to design algorithms that are interpretable, and preferably, theoretically sound.

In general, my interests lie in the mathematical foundations of computer science - information theory, linear algebra, graph theory and combinatorics to name a few. I also picked up a curiosity for functional programming and distributed systems during my stint as a software engineer at Informatica.

Selected Publications:

1. Abhra Chaudhuri, Massimiliano Mancini, Zeynep Akata and Anjan Dutta. "Relational Proxies: Fine-Grained Relationships as Zero-Shot Discriminators". In IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI), 2024.
2. Abhra Chaudhuri, Serban Georgescu, and Anjan Dutta. "Learning Conditional Invariances through Non-Commutativity". In the Proceedings of International Conference on Learning Representations (ICLR), Vienna, Austria, 2024.
3. Abhra Chaudhuri, Massimiliano Mancini, Zeynep Akata and Anjan Dutta. "Transitivity Recovering Decompositions: Interpretable and Robust Fine-Grained Relationships". In the Proceedings of Neural Information Processing Systems (NeurIPS), New Orleans, USA, 2023.
4. Abhra Chaudhuri, Ayan Kumar Bhunia, Yi-Zhe Song and Anjan Dutta. "Data-Free Sketch-Based Image Retrieval". In the Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, Canada, June, 2023.
5. Abhra Chaudhuri, Massimiliano Mancini, Zeynep Akata and Anjan Dutta. "Relational Proxies: Emergent Relationships as Fine-Grained Discriminators". In the Proceedings of Neural Information Processing Systems (NeurIPS), New Orleans, USA, 2022 (Spotlight).
6. Abhra Chaudhuri, Massimiliano Mancini, Yanbei Chen, Zeynep Akata and Anjan Dutta. "Cross-Modal Fusion Distillation for Fine-Grained Sketch-Based Image Retrieval". In the Proceedings of British Machine Vision Conference (BMVC), London, UK, 2022.