Kavya Gupta

Postdoctoral Researcher at Saarland University with Prof. Isabel Valera, working on Society-Aware Machine learning.

Ph.D. Graduate specializing in the Robustness of neural networks from Ecole CentraleSupelec, University of Paris-Saclay and funded by Thales LAS France. I was advised by Prof. Jean-Christophe Pesquet. My thesis revolves around Lipschitz Constant Analysis of Neural Networks for certified Robustness and adversarial machine learning. I am interested in the Robustness, trustworthiness, and reliability of neural networks.

Before moving to France, I worked at Tata Consultancy Services Research and Innovation (2016-2019), Kolkata, India. I worked in Robotics and Vision Group. I worked on 2D and 3D human pose estimation from monocular videos, object detection and classification, inverse problems for removing blur in images captured from fast-moving drones, and rendering personalized realistic facial animation of an avatar using 3D Face meshes synced with audio.

I completed my M.Tech in Electronics and Communication Engineering with specialization in Communication and Signal Processing from IIIT-Delhi and was a member of SALSA Research Lab (Batch 2014-2016). I did my master’s thesis on Regularized Autoencoders under the guidance of Dr. Angshul Majumdar. I also worked on compressive sensing and collaborative filtering.

Research Interests

• Society-Centered AI, Robust and Interpretable AI

• Causality-driven Decision-Making

• Multi-objective Optimization

• Formal guarantees of robustness of neural networks