Kaustubh Mani

I am an incoming Ph.D. student at the at Mila and the Robotics and Embodied AI Lab , advised by Prof. Liam Paull. My research goal is to design safe and data-efficient learning algorithms with emphasis on real-world robotics implementations.

I received my master's degree at Robotics Research Center, IIIT Hyderabad working with Prof. K. Madhava Krishna on scene understanding for autonomous driving. I did my undergraduate studies at Indian Institute of Technology, Kharagpur where I worked with Prof. Pabitra Mitra

Twitter / GitHub / Google Scholar

Research

	f-Cal: Calibrated aleatoric uncertainty estimation from neural networks for robot perception Dhaivat Bhatt, Kaustubh Mani, Dishank Bansal, J. Krishna Murthy, Liam Paull arXiv 2021 (under submission) Learning to predict calibrated uncertainty estimates for robotic perception tasks. #Uncertainty_Calibration #BayesianNN #Robotic_Perception [Paper][Website]
	AutoLay: Benchmarking Monocular Layout Estimation Kaustubh Mani, N. Sai Shankar, J. Krishna Murthy, K. Madhava Krishna IROS, 2020 A new dataset and a benchmark for amodal layout estimation from monocular imagery. [Paper] [Code] [Website]
	MonoLayout: Amodal scene layout from a single image Kaustubh Mani, Swapnil Daga, Shubhika Garg, N. Sai Shankar, J. Krishna Murthy, K. Madhava Krishna WACV, 2020 Learning to "hallucinate" layout of a road scene from a single monocular image, including parts of the image that are occluded/partly-visible. [Paper] [Code] [Website]
	Learning Adaptive driving behavior using Recurrent Deterministic Policy Gradients Kaustubh Mani, Meha Kaushik, Nirvan Singhania, K. Madhava Krishna NeurIPS 2019 Workshop on Machine Learning for Autonomous Driving(ML4AD) Improving adaptibility in varying traffic scenarios. #Model_Free_RL [Paper]

Source