.
├── data
│ ├── 0_dB_pump.zip
│ └── pump
│ ├── id_00
│ │ ├── abnormal
│ │ └── normal
│ ├── id_02
│ │ ├── abnormal
│ │ └── normal
│ ├── id_04
│ │ ├── abnormal
│ │ └── normal
│ └── id_06
│ ├── abnormal
│ └── normal
├── makefile
├── nn.py
├── README.md
├── requirements.txt
└── tools
├── __init__.py
├── prepro.py
└── show.py
- Tools dir for import tools and feature extracting tools
- datadir for the datafiles
- makefile for downloading audio data and cleaning and so on
- requirements.txt for managing requirements. Good to use with virtualenv
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
normalization (Normalization (None, 1598, 20, 1) 3
_________________________________________________________________
conv2d (Conv2D) (None, 1596, 18, 60) 600
_________________________________________________________________
conv2d_1 (Conv2D) (None, 1594, 16, 60) 32460
_________________________________________________________________
flatten (Flatten) (None, 1530240) 0
_________________________________________________________________
dense (Dense) (None, 60) 91814460
_________________________________________________________________
dense_1 (Dense) (None, 2) 122
=================================================================
Total params: 91,847,645
Trainable params: 91,847,642
Non-trainable params: 3
_________________________________________________________________
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
resizing (Resizing) (None, 32, 32, 1) 0
_________________________________________________________________
normalization (Normalization (None, 32, 32, 1) 3
_________________________________________________________________
conv2d (Conv2D) (None, 30, 30, 60) 600
_________________________________________________________________
flatten (Flatten) (None, 54000) 0
_________________________________________________________________
dense (Dense) (None, 60) 3240060
_________________________________________________________________
dense_1 (Dense) (None, 2) 122
=================================================================
Total params: 3,240,785
Trainable params: 3,240,782
Non-trainable params: 3
_________________________________________________________________