Avinash Kori

Avinash Kori is a Ph.D. researcher at Imperial College London affiliated with the Safe and Trusted AI Centre for Doctoral Training (CDT). Supervised by Prof. Francesca Toni and Prof. Ben Glocker, his research focuses on Explainable AI (XAI), causality, and deep learning with applications in medical image analysis and optimization algorithms.

His work includes publications on arXiv and conferences like MICCAI, covering topics such as robust segmentation, concept-based explanations, and symbolic reasoning in hyperbolic space. He has also explored stochastic optimization, support vector machines (SVM), and gradient descent variants, providing theoretical and practical implementations.

Recent trends in his publications highlight advancements in robust CNN models, causal logic frameworks, and hyperbolic geometry for hierarchical learning. His research is driven by the need to make AI systems more transparent and reliable for critical domains like healthcare.

Scientific Awards:

• AAAIw Overall Best Paper Award (Feb 2021) for CNN interpretability research.

He actively contributes to open-source implementations via platforms like GitHub and shares insights through blogs and paper reviews. His academic journey includes an undergraduate degree in Biomedical Engineering Design with a minor in Machine Learning from Indian Institute of Technology, Madras, followed by research internships at Siemens and Stanford University.