innvestigate.applications

Imagenet

Example applications for image classifcation.

Each function returns a pretrained ImageNet model. The models are based on keras.applications models and contain additionally pretrained patterns.

The returned dictionary contains the following keys: model, in, sm_out, out, image_shape, color_coding, preprocess_f, patterns.

Function parameters:

param load_weights:  Download or access cached weights. param load_patterns:  Download or access cached patterns.

innvestigate.applications.imagenet.vgg16(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.vgg19(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.resnet50(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.inception_v3(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.inception_resnet_v2(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.densenet121(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.densenet169(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.densenet201(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.nasnet_large(load_weights=False, load_patterns=False)

innvestigate.applications.imagenet.nasnet_mobile(load_weights=False, load_patterns=False)