Architecture | Convolution, 1x1 Convolution, Dense Connections, Depthwise Separable Convolution, ReLU, Global Average Pooling, Max Pooling, Softmax, Residual Connection |
---|---|
ID | xception |
SHOW MORE |
Architecture | Convolution, 1x1 Convolution, Dense Connections, Depthwise Separable Convolution, ReLU, Global Average Pooling, Max Pooling, Softmax, Residual Connection |
---|---|
ID | xception41 |
SHOW MORE |
Architecture | Convolution, 1x1 Convolution, Dense Connections, Depthwise Separable Convolution, ReLU, Global Average Pooling, Max Pooling, Softmax, Residual Connection |
---|---|
ID | xception65 |
SHOW MORE |
Architecture | Convolution, 1x1 Convolution, Dense Connections, Depthwise Separable Convolution, ReLU, Global Average Pooling, Max Pooling, Softmax, Residual Connection |
---|---|
ID | xception71 |
SHOW MORE |
Xception is a convolutional neural network architecture that relies solely on depthwise separable convolution layers.
To load a pretrained model:
import timm
m = timm.create_model('xception', pretrained=True)
m.eval()
Replace the model name with the variant you want to use, e.g. xception
. You can find the IDs in the model summaries at the top of this page.
You can follow the timm recipe scripts for training a new model afresh.
@article{DBLP:journals/corr/ZagoruykoK16,
@misc{chollet2017xception,
title={Xception: Deep Learning with Depthwise Separable Convolutions},
author={François Chollet},
year={2017},
eprint={1610.02357},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
MODEL | TOP 1 ACCURACY | TOP 5 ACCURACY |
---|---|---|
xception71 | 79.88% | 94.93% |
xception65 | 79.55% | 94.66% |
xception | 79.05% | 94.4% |
xception41 | 78.54% | 94.28% |