F1 Steering Angle Model

A machine learning application that predicts steering angles from F1 onboard camera footage.

About the Project

Steering input is one of the key fundamental insights into driving behavior, performance and style. However, there is no straightforward public source, tool or API to access steering angle data. The only available source is onboard camera footage, which comes with its own limitations, such as camera position, shadows, weather conditions, and lighting.

• The F1 Steering Angle Prediction Model uses a CNN based on EfficientNet-B0 to predict steering angles from a F1 onboard camera footage, trained with over 25,000 images (7000 manual labaled augmented to 25000) and YOLOv8-seg nano for helmets segmentation, allowing the model to be more robust by erasing helmet designs.

• Currentlly the model is able to predict steering angles from 180° to -180° with a 3°-5° of error on ideal contitions.

• EfficientNet-B0 and YOLOv8-seg nano are exported to ONNX format, and images are resized to 224x224 allowing it to run on low-end devices.

How It Works

Video Processing:

• From the onboard camera video, the frames selected are extracted at the FPS rate.

Image Preprocessing:

• The frames are cropeed based on selected crop type to focus on the steering wheel and driver area.
• YOLOv8-seg nano is applied to the cropped images to segment the helmet, removing designs and logos.
• Convert cropped images to grayscale and apply CLAHE to enhance visibility.
• Apply adaptive Canny edge detection to extract edges, helped with preprocessing techniques like bilateralFilter and morphological transformations.

Prediction:

• The CNN model processes the edge image to predict the steering angle

Postprocessing

• apply local a trend-based outlier correction algorithm to detect and correct outliers

Results Visualization

• Angles are displayed as a line chart with statistical analysis also a csv file with the frame number, time and the steering angle.

Original Onboard Frame

Preprocessed Images

Left to right: Cropped image, Yolo segmentation, CLAHE enhanced image, Edge detection result

Performance Analysis

After extensive development, the model has achieved the following performance metrics:

• From 0 to ±90° = 3° of ground truth
• From ±90° to ±180° = 5° of ground truth

Limitations: Performance may decrease in:

• Low visibility conditions (rain, extreme shadows)
• Low quality videos (low resolution, high compression)
• Changed camera positions (different angle, height)

Installation and Usage

Prerequisites

• Python 3.11
• requirements.txt dependencies

Option 1: Use the Hosted Version

The simplest way to use the application is through the hosted Streamlit app.

Option 2: Run Locally

1. Clone the repository

   git clone https://github.com/danielsaed/F1-machine-learning-webapp.git
   cd F1-machine-learning-webapp

2. Install dependencies

   # For Linux/macOS, install OpenCV system dependencies
   sudo apt update
   sudo apt install -y libgl1-mesa-glx
   
   # Install Python dependencies
   pip install -r requirements.txt

3. Run the application

   streamlit run streamlit_app.py

4. Open in browser
   The application will be available at http://localhost:8501

Option 3: Download the Standalone App (Windows)

1. Go to the Releases page of this repository.
2. Download the latest .exe file (e.g., F1SteeringApp.exe) from the Assets section.
3. (Optional) Place your .env file in the same folder as the .exe if you need custom environment variables.
4. Double-click the .exe to launch the app.
5. The app will open in your default browser.

---

Option 4: Run with Docker

a) Using the pre-built Docker image

1. Install Docker if you don't have it.
2. Run the following command:

   docker run -p 8501:8501 danielsaed99/f1-steering-angle

3. Open your browser at http://localhost:8501

b) Build and run from the Dockerfile

1. Clone the repository

   git clone https://github.com/danielsaed/F1-machine-learning-webapp.git
   cd F1-machine-learning-webapp

2. Build the Docker image

   docker build -t f1-steering-angle .

3. Run the container

   docker run -p 8501:8501 f1-steering-angle

4. Open your browser at http://localhost:8501

Contributing

Any contributions to improve the model or application are welcome.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Disclaimer

Downloading or recording F1 onboards videos potentially violates F1/F1TV's terms of service.

This model is for research/educational purposes only. It is not related to F1 or any other organization.