[ECCV 2024 Oral] DriveLM: Driving with Graph Visual Question Answering

End to End Autopilot Perception Playbook

In this project, I used Python and TensorFlow to classify traffic signs. Dataset used: German Traffic Sign Dataset. This dataset has more than 50,000 images of 43 classes. I was able to reach a +99% validation accuracy, and a 97.3% testing accuracy.

📐 Personal GitHub web page. Based on the minimal-mistakes Jekyll theme.

Multi-Object Tracking and Trajectory Prediction. This repository contains all codes written for SUMMER RESEARCH INTERNSHIP (2021) at AI and Robotics Park (ARTPARK), IISc Bangalore

Countdown for all* relevant conferences in the domain of autonomous driving

Inspired by comma.ai OpenPilot idea, this is an AD Autopilot on RaspberryPi based on ROS called RosPilot. The project currently performs lane follow feature. The codebase is written to be modular, enable quick prototyping and facilitate learning and collaboration across multiple users. The hardware used so far is the donkey car robocar + RPI 3

Lokale Navigation von Mikromobilitätsfahrzeugen mittels Reinforcement Learning

My personal website

This repository hosts the Zenseact Open Dataset website.

Jekyll code for the website Computationally Thinking Blog

Computer vision project on vehicle detection and tracking on roads

A framework for the analysis of trust in the interaction between pedestrians and vehicle (manual and automated), from the perspective of the driver of a manual or an automated vehicle, using a crowdsourcing approach.

This is a Deep Learning Project to classify German Road Signs using deep neural networks and image processing.

A framework for the analysis of un(certainty) in traffic, using a crowdsourcing approach.

Project on detecting lane lines on roads in a video stream, using polynomials

Application to detect and mark lane lines on the road.

Udacity Introduction to self-driving cars Nanodegree

[Small] Traffic sign classification using Tensorflow and LeNet.