Autonomous Navigation using Optical flow and its uncertainty

A pipeline to combine the data from LiDAR and Camera to obtain semantic painted point cloud. 

Object detection after filtering, segmentation and clustering of point cloud data.

3D reconstruction (sparse) and simultaneous localization of a monocular camera.

Panorama image stitching using classical and Deep learning computer vision methods.

End-to-end pipeline to swap faces in a video using facial landmarks correspondences between faces.

Synthesize novel views of scenes by optimizing a continuous volumetric scene function using a sparse set of input views.

Zhang's camera calibration method implementation to estimate the camera intrinsics and distortion parameters

Trained a deep learning model to predict steering angles so that the vehicle can navigate autonomously in a Unity simulator environment.

Live video-feed processing pipeline to detect lane markings among other terrestrial features

Implementation of path planning algorithms like Astar, BFS, DFS, Dijkstra on an obstacle grid.