Skip to content

Artificial Intelligence · Capstone

Self-Driving Car Simulator with Behavioural Cloning

A CNN learns to steer by watching you drive in a simulator

4-5 weeks build

Overview

You drive laps in the open-source Udacity driving simulator to record camera frames and steering angles, then train an NVIDIA-style CNN to predict steering from a single image. A socket server feeds the trained model back into the simulator for autonomous laps, with a lane-detection overlay drawn on the live feed to show what the road looks like to the car.

What makes it stand up

  • 01Data collection guide for balanced, recoverable driving laps
  • 02Augmentation with flips, shifts, brightness and side-camera offsets
  • 03End-to-end CNN regressing steering angle from one frame
  • 04Live autonomous mode driving the simulator through a socket bridge
  • 05Lane overlay and steering trace plotted during each run

Module breakdown

  1. 01

    Recording

    Simulator capture, CSV logs and steering histogram balancing.

  2. 02

    Augmentation

    Image transforms that teach recovery from road edges.

  3. 03

    Model

    Convolutional regressor with training curves and checkpoints.

  4. 04

    Drive server

    Flask-SocketIO bridge sending steering and throttle live.

  5. 05

    Overlay

    Lane detection drawn over the simulator feed.

After this, you will be able to

  • Explain behavioural cloning and its distribution-shift problem
  • Defend your augmentation choices with before-and-after runs
  • Demo the car completing a lap on its own

Technology stack

  • Python
  • TensorFlow
  • OpenCV
  • Flask
  • NumPy

Complete kit

₹8,999₹15,499
-42%

Delivered in under 48 hours

Level
Capstone
Domain
Artificial Intelligence
Build time
4-5 weeks
Difficulty

You receive

  • Complete, commented source code
  • 60-90 page project report (IEEE format)
  • Editable presentation deck
  • Architecture, ER and UML diagrams
  • Local setup and deployment guide
  • Sample dataset or seed data
  • Viva question bank with answers
  • One 45-minute walkthrough call
  • 30 days of doubt-clearing support
Compare tiers

Keep looking

Students also considered

PopularMajor Project

Artificial Intelligence

Face Recognition Attendance System

Camera-based attendance with liveness detection and analytics

  • Python
  • OpenCV
  • FaceNet
  • Flask
  • +2
₹4,900₹5,999
4.8/301 reviews
PopularMajor Project

Artificial Intelligence

Mental Health Support Chatbot

Grounded conversational agent with sentiment tracking

  • Python
  • Transformers
  • RAG
  • FastAPI
  • +2
₹4,699₹7,999
4.8/143 reviews
Major Project

Artificial Intelligence

Traffic Sign Recognition

Real-time detection and classification for driver assistance

  • Python
  • YOLOv8
  • OpenCV
  • PyTorch
  • +2
₹3,899₹6,799
4.7/126 reviews