This Capsulation Library consists of Deep Learning with PyTorch and some signal processing with scipy(or others).
-
data
- /dataset/ is where you store your data
- /my*.py provides an example realization of torch.dataloader
-
models
- /few-shot/ provides some networks about few-shot learning
- /machine-learning/ provides some machine learning algorithms, including classification, regression and clusters.
- /*.py provides some classical network architechtures.
-
signal-processing
- /features/ provides time、frequency、energy domain features extraction.
- /filters/ provides some signal filters.
- /transformation/ provides some signal transformation (time->freq or freq->time, etc)
-
utils
- /pytorch2mobile provides some converter on how to convert PyTorch model to ncnn or TorchMobile
- /*.py provides some tools to handle time calculation, etc.
-
visualization
- /*.py provides some visualization functions.