urjc-deepracer/carla-expert-racing

URJC-DeepRacer: Autonomous Driving Dataset

This dataset was generated by Sergio Robledo as part of the URJC-DeepRacer project, focused on training and validating autonomous driving agents using Deep Learning and Reinforcement Learning techniques.

Data was collected using the CARLA Simulator, featuring DeepRacer vehicle models and custom-designed racing environments.

📊 Dataset Overview

Each entry provides a synchronized capture of the front-facing RGB camera, its corresponding semantic segmentation mask, and the vehicle's real-time telemetry.

Features

• file_name: RGB Image (Front-facing vehicle camera).
• mask_path: Semantic Segmentation Mask (Ground truth for vision tasks).
• speed: Current vehicle speed (m/s).
• steer: Steering angle (normalized between -1 and 1).
• throttle: Throttle intensity (0 to 1).
• brake: Brake intensity.
• heading: Vehicle orientation.
• experiment_id: Unique identifier for the simulation session/map.
• maneuver: Agent state indicator during collection.

🚀 Getting Started

You can load this dataset directly using the Hugging Face datasets library:

from datasets import load_dataset

# Load the dataset from the URJC-DeepRacer organization
dataset = load_dataset("urjc-deepracer/your-repo-name")

# Access the first sample of the training split
sample = dataset['train'][0]

# Display images
sample['file_name'].show()
sample['mask_path'].show()

print(f"Speed: {sample['speed']} m/s | Steer: {sample['steer']}")

🛠️ Data Collection Methodology

The data is organized by individual experiments to facilitate the study of various scenarios, including different weather conditions, maps, and traffic densities. Using the jsonl format ensures efficient and scalable data streaming.

🎓 Credits & Affiliation

This dataset is maintained by URJC DeepRacer developers as part of ongoing autonomous driving research.