Sparsh913

Learning-for-3D-Vision

Visual-Learning-Recognition

16-824 Assignments

VLR-Project

Solving robotic manipulation through vision and other inputs

RAFT-Retrain-through-NeRF

Using NeRF representation to retrain a deep stereo network, RAFT to make it perform better on Riesling vines dataset

torch-tutorials

Coding up basic torch functionalities to learn the best practices in PyTorch

sparshgarg

Diffusion-Models

Coding a Diffusion model from scratch

3D-vines

3D Reconstruction of vines from stereo images

Deep-Learning-Temp

3D-Reconstruction

Using Geometry to do 3D reconstruction from 2 images of a temple

24677-Modern_Control_Theory

Tesla Model 3 Control

Lukas-Kanade-Tracking

Building Lucas Kanade Tracking Algorithm from ground up

Augmented-Reality-Planar-Homography

Scene-Classification-using-Image-Processing

Undoubtedly, basic ML models can classify a scene as a building, a park, a kitchen and so on. But in this project, I tried to do it without any learning, applying the basic principles of Image Processing.