Supervisors: Dr. Krikamol Muandet (MPI-IS) and Prof. Bernhard Schölkopf (MPI-IS/ETHz)

Robustness-enhancement vs Bias

Collaborators: Prof. Tatsu Hashimoto (Stanford) and Prof. Narendra Ahuja (UIUC); Ruchit Rawal.

Estimating drought-resilience and livestock sustenance through satellite (multi-spectral) images, to enhance public policy and planning. Collaborators: Ruchit Rawal

Compressed NNs for SoTA image-classification. (Team Award: Google Cloud Credits x25 TPU v2)

Advisor: Prof. Narendra Ahuja (UIUC) | Collaborators: Gopi Kishan, Meenakshi Sarkar

As a part of #AIforSocialGood, I am working on designing a segmentation network for unmanned systems deployed in crisis response. Special focus is on robustness with domain-adaptation pipeline and reliable safety. Collaborators: Sachin Shriwastav, Meenakshi Sarkar

As a member of the computer vision group, I worked on developing a computationally light and reliable visual-servoing platform for the heterogeneous vehicle system (UAV+UGV).

*Finals* : IISc-TCS team won the "Best Innovation" Prize and presented reports.

Mentors: Dr. Kaushik Das (TCS R&D Lab), Prof. Raghu Krishnapuram (RBCCPS)

We've created a robot-motion dataset using LEGO bots & Turtlebot to validate various motion prediction or other deep networks. Currently, we're working to include even more agents, complex scenes, trajectories, and plan to expand it to UAVs. Collaborators: Meenakshi Sarkar

Study of hardware-related vulnerabilities of remote-controlled robots.

Wind tunnel helps in fine-tuning the design of aircrafts. Models are tested to scale. Suction-type design and Smoke (Flow) Visualization of airfoils.