About Us
We develop state-of-the-art algorithms for image and video enhancement, restoration, and generative modeling. Our focus is on theoretical rigor, real-time efficiency, and practical deployability across a variety of imaging tasks.
Research Areas
Learning-based Iterative Algorithms
Provably convergent and certifiable learning-based algorithms for restoration tasks.
Real-time Image and Video Restoration
Efficient state-of-the-art models deployable on edge devices.
Video Stabilization
Motion-aware stabilization for high-motion and gyroscope-informed scenarios.
Deepfake Video Detection
Robust detection methods generalizing to diffusion-based deepfakes.
Generative Models
Developing text-to-image generative models for Indian content.
Real-time Generative Models
Develop mathematical models for real-time flow and diffusion models.