GVS Mothish

PhD Student · Computational and Data Sciences, IISc Bangalore

2025 Google PhD Fellow · Machine Perception

I'm a third-year PhD student at CDS, IISc Bangalore, advised by Dr. Deepak Subramani. I work on generative models, computer vision, neural operators and reasoning in foundation models.

Before my PhD, I did an MTech in Robotics and Autonomous Systems at RBCCPS, IISc, with Dr. Shishir N. Y. at the Stochastic Robotics Lab, building RL- and diffusion-based locomotion for walking robots. I'm broadly interested in intelligent stochastic systems.

Email CV GitHub LinkedIn Twitter
GVS Mothish

Research interests

Publications · * equal contribution

CN-Diff diagram
ICML 2025

Conditional Diffusion Model with Nonlinear Data Transformation for Time Series Forecasting

J Rishi*, GVS Mothish*, Deepak Subramani

CN-Diff: nonlinear transformations and learnable conditions in the forward diffusion process, with a new training loss and full derivations of the forward and reverse processes.

PaperCode
DNOD
ICLR 2026

DNOD: Deformable Neural Operators for Object Detection in SAR Images

GVS Mothish, J Rishi, Shobhit Kumar Shukla, Deepak Subramani

A neural-operator detector for SAR imagery with Multi-Scale Fourier Mixing in the encoder and an adaptive deformable Fourier neural operator in the decoder; +2.23 mAP on SARDet-100k.

PaperCode
IEEE ICC 2024

BiRoDiff: Diffusion Policies for Bipedal Robot Locomotion on Unseen Terrains

GVS Mothish, Manan Tayal, Shishir Kolathaya

A real-time diffusion-policy controller trained offline that captures multiple gaits and velocities in one policy and generalizes to unseen terrain.

ProjectPaperVideo
Stoch BiRo robot
IEEE ICCAR 2024

Stoch BiRo: Design and Control of a Low-cost Bipedal Robot

GVS Mothish, Karthik Rajgopal, Ravi Kola, Manan Tayal, Shishir Kolathaya

A modular, point-footed bipedal robot with proprioceptive actuation and a linear-policy walking controller running on a Raspberry Pi 4.

ProjectPaperVideo
Human-robot handover setup
ACM IUI 2024

Multimodal Target Prediction for Rapid Human-Robot Interaction

Mukund Mitra, Ameya Avinash Patil, GVS Mothish, Gyanig Kumar, Abhishek Mukhopadhyay, Murthy LRD, Partha Pratim Chakraborty, Pradipta Biswas

Bayesian fusion of eye gaze and hand motion, with inverse RL to learn human preferences for human-robot handover.

Paper

Awards & honors