AutoDrive-Vision

Multi-Model approach for Autonomous Driving using Deep Learning and OpenCv

• Tech Stack: Python,Tensorflow, PyTorch, OpenCv, Pillow, Keras, Numpy, Pandas, Jetson Nano, Arduino
• Github URL: Project Link

Developed Multiple Deep Learning Models to detect and classify Traffic signals, detect obstacles and detect lanes.

Performed behavioural cloning of self-driving cars in simulated environments using neural networks.

Performed comparative study with different models like Mask-RCNN, ResNet50, InceptionV3 and MobileNet.

Worked on KITTI 3D data visualization, FCNN, DeepSort, MTAN, SFA 3D, UNetXST algorithms.

Built an Autonomous driving vehicle using Jetson Nano, Arduino, Ultrasonic Sensor that can perform Lane Detection, Obstacle Avoidance and response to Traffic Signals using Deep Learning and Image Segmentation.

Designed a self-driving reinforcement learning model using Deep Q-learning in OpenAI Gym environment.