This project trains a TensorFlow model in SageMaker for traffic sign recognition.
The files contained in this repo includes:
- readme.md
- Traffic_sign_recognition_with_TensorFlow.ipynb
- train.py
Run Traffic_sign_recognition_with_TensorFlow.ipynb in SageMaker as a notebook instance.
Necessary files not included in this repo can be downloaded from Kaggle: GTSRB - German Traffic Sign Recognition Benchmark.
Here is an overview of the dataset:
- Each image has slightly different sizes
- Number of training examples = 39209
- Number of testing examples = 12630
- Number of classes = 43
The network architecture is similar to LeNet and the details of each layer is as follows:
Layer | Description |
---|---|
Input | 32x32x1 Grayscale image |
Convolution 5x5 | 1x1 stride, valid padding, outputs 28x28x6 |
RELU | |
Max pooling | 2x2 stride, outputs 14x14x6 |
Convolution 5x5 | 1x1 stride, valid padding, outputs 10x10x16 |
RELU | |
Max pooling | 2x2 stride, outputs 5x5x6 |
Fully connected | Input = 400. Output = 120. |
Fully connected | Input = 120. Output = 84. |
Dropout | |
Fully connected | Input = 84. Output = 43. |
Softmax | Output layer |
See complete tutorial here