Deep Learning and Computer Vision 3D Scene Reconstruction from Monocular Images using SFM A classical computer vision pipeline for 3D scene reconstruction using the Structure from Motion (SfM) algorithm. Diffusion Models for Image Data Generation An application of Denoising Diffusion Probabilistic Models (DDPM) to generate high-quality synthetic training data for an image classifier using the CIFAR-10 dataset. Adversarial Patch Generation for Monocular Depth Networks Customised implementation of this paper by Yamanaka et. al. to attack Depth networks with Adversarial patches. Semantic and Instance Segmentation of Aerial Drone Imagery A custom CNN model inspired by ResNet-18 that can perform Semantics on footage obtained from drones. Robot Control and Navigation High to Low Level Task Planning and Execution using RL An RL framework for mobile manipulators to learn navigation and grasping tasks sequentially for Household tasks Quadrotor Control and Trajectory Optimization Tracking complex trajectories of Quadrotors using PID and optimal controllers like LQR and MPC Navigation of a Differential Drive Robot with LIDAR based EKF SLAM A short implementation of EKF SLAM for a robot to navigate in a maze like environment
