Open pribadihcr opened 7 years ago
Hello, To extract the CNN feature you can execute the saveFeats.py, entering the configuration file, the list of sample inputs (images), the window size (for sliding window step) and the output folder, where the features will be saved. You can acquire the pretrained models on
VGG-Places http://places.csail.mit.edu/downloadCNN.html
VGG-Imagenet (ILSVRC-2014) http://www.robots.ox.ac.uk/~vgg/research/very_deep/
Also, we will offer a detailed documentation soon.
hi,when will we see a detailed documentation?thanks
Hi,
Is there readme tutorial how to get CNN feature and the pretrained model ?
Thanks