-
Notifications
You must be signed in to change notification settings - Fork 19
Models
In total, neural-dream supports 26 different models. All of the supported models can be found at the links below:
-
All of the models listed on the neural-style-pt Wiki.
-
All of the pre-trained GoogleNet, Inception, and VGG models listed by Torchvision.
-
The MIT Places205 and Places365 GoogleNet models from Caffe, converted to PyTorch here.
-
The GoogleNet SOS, GoogleNet cars, BVLC GoogleNet, and ResNet Open NSFW models converted from Caffe to PyTorch here.
-
The Inception5h (inception v1) model from TensorFlow and the Keras Inception V3 model, converted to PyTorch here.
The items listed under "Download Option Name" can be used to download specific models with neural-dream's included download_models.py
script. To download the GoogleNet MIT Places365 model for example, you would use:
# Replace models/download_models.py with neural-dream if using pip version
python models/download_models.py -models caffe-googlenet-places365
- neural-dream will automatically select the right preprocessing/deprocessing, input range, and input mean values for all the models listed below.
Name | Filename | Download Option Name |
---|---|---|
VGG-19 ILSVRC | vgg19-d01eb7cb.pth |
caffe-vgg19 |
VGG-16 ILSVRC | vgg16-00b39a1b.pth |
caffe-vgg16 |
VGG-16 Salient Object Detection (SOD) Finetune | VGG16_SOD_finetune.pth |
None |
VGG-16 PASCAL VOC FCN-32s | fcn32s-heavy-pascal.pth |
None |
VGG-16 PASCAL VOC NYUD FCN-32s Color Heavy | nyud-fcn32s-color-heavy.pth |
None |
VGG-16 FCN-32s PASCAL-Context | pascalcontext-fcn32s-heavy.pth |
None |
VGG-16 FCN-32s SIFT Flow | siftflow-fcn32s-heavy.pth |
None |
VGG-16 Channel Pruning | channel_pruning.pth |
None |
VGG-16 MIT Places 365 | vgg16_places365.pth |
None |
VGG-16 MIT Places Hybrid 1365 | vgg16_hybrid1365.pth |
None |
VGG-16 Stylized-ImageNet | VGG16-Stylized-ImageNet.pth |
None |
PyTorch VGG-11 | vgg11-bbd30ac9.pth |
None |
PyTorch VGG-13 | vgg13-c768596a.pth |
None |
PyTorch VGG-16 | vgg16-397923af.pth |
pytorch-vgg16 |
PyTorch VGG-19 | vgg19-dcbb9e9d.pth |
pytorch-vgg19 |
Name | Filename | Download Option Name |
---|---|---|
BVLC GoogleNet | bvlc_googlenet.pth |
caffe-googlenet-bvlc |
GoogleNet MIT Places205 | googlenet_places205.pth |
caffe-googlenet-places205 |
GoogleNet MIT Places365 | googlenet_places365.pth |
caffe-googlenet-places365 |
GoogleNet Cars | googlenet_finetune_web_cars.pth |
caffe-googlenet-cars |
GoogleNet SOS | GoogleNet_SOS.pth |
caffe-googlenet-sos |
PyTorch GoogleNet | googlenet-1378be20.pth |
pytorch-googlenet |
Name | Filename | Download Option Name |
---|---|---|
TensorFlow Inception5h | inception5h.pth |
tensorflow-inception5h |
Pytorch Inception V3 | inception_v3_google-1a9a5a14.pth |
pytorch-inceptionv3 |
Keras Inception V3 | inceptionv3_keras.pth |
keras-inceptionv3 |
Name | Filename | Download Option Name |
---|---|---|
ResNet Yahoo Open NSFW | ResNet_50_1by2_nsfw.pth |
caffe-resnet-opennsfw |
Name | Filename | Download Option Name |
---|---|---|
NIN Imagenet | nin_imagenet.pth |
caffe-nin |