Skip to content

Implementation of "CNN-Based Joint Clustering and Representation Learning with Feature Drift Compensation for Large-Scale Image Data"

License

Notifications You must be signed in to change notification settings

kapitsa2811/cnn-clustering

Repository files navigation

Pretrained AlexNet model for TensorFlow

This repository is a fork of kratzert/finetune_alexnet_with_tensorflow, and have been adapted to generate a frozen protobuf for AlexNet.

First download alexnet weights (from caffee) in .npy format:

wget http://www.cs.toronto.edu/~guerzhoy/tf_alexnet/bvlc_alexnet.npy

Put the weights into the same directory as the this git repository.

Then run dump_pb.py

python3 dump_pb.py

Then a network with trainable weights is saved to alexnet.pb, and a frozen protobuf is saved to alexnex_frozen.pb

Download saved alexnet.pb and alexnet_frozen.pb here:

alexnet.pb

alexnet_frozen.pb

About

Implementation of "CNN-Based Joint Clustering and Representation Learning with Feature Drift Compensation for Large-Scale Image Data"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published