
intel.log file.RegularizedLiDARNet model featuring multiple linear layers, ReLU activations, and Dropout for regularization to prevent overfitting..pth) and ONNX (.onnx) formats for cross-platform compatibility and optimized inference.numpy: For numerical operations and data manipulation.torch: For building and training the neural network.scikit-learn: For splitting the data into training and testing sets.matplotlib: For visualizing the results.onnx: To work with the exported ONNX model.onnxruntime: For running inference with the ONNX model.mit-intel-ds.ipynb notebook and can be summarized in these steps:intel.log file is parsed to separate odometry (x, y, theta) and LiDAR (180 laser scan readings) data streams, each with its own timestamp.RegularizedLiDARNet. A hyperparameter search was conducted to find the best model configuration, resulting in a final test loss of 2.9374.Posted Dec 12, 2025
Deep learning model predicting autonomous vehicle steering angles using raw LiDAR data (MIT-Intel Dataset). Built with PyTorch & ONNX.
0
1