Sovereign 8s.is mirror of standardgalactic/NEOSTI
- Python 100%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
|
|
||
| code | ||
| data | ||
| LICENSE | ||
| README.md | ||
| readme.md | ||
NEOSTI - A Neuromorphic Electronic-Opto Spatial-Temporal Hybrid Image Sensor
This is a sample code for training NEOSTI model on a demo dataset, UnityEyes dataset.
Requirements
- Python 3.7+
- numpy >= 1.21.0
- torch >= 1.1.0
- torchvision >= 0.3.0
- pillow >= 9.1.0
- scipy >= 1.3.0
Instructions
- Enter the ./code directory.
- Run the demo script main.py.
python main.py
Output
The demo script is expected to:
- Print the loss and accuracy during the training process.
- Save the current optimal model parameters in
./results/best_model.pthduring the training process. - Print the accuracy and convolution kernel parameters of the best model after training.
License
This project is licensed under the MIT License.