Convenient wrapper for TensorFlow feature extraction from pre-trained models using tf.contrib.slim
Switch branches/tags
Nothing to show
Clone or download
Latest commit 128e8a9 May 9, 2018

TensorFlow Feature Extractor

This is a convenient wrapper for feature extraction or classification in TensorFlow. Given well known pre-trained models on ImageNet, the extractor runs over a list or directory of images. Optionally, features can be saved as HDF5 file. It supports all the pre-trained models listed on the official page.

TensorFlow models tested:

  1. Inception v1-v4
  2. ResNet v1 and v2
  3. VGG 16-19


  • TensorFlow (tested with version 1.8)
  • TensorFlow Models
  • The usual suspects: numpy, scipy.
  • Optionally h5py for saving features to HDF5 file


  1. Checkout the TensorFlow models repository somewhere on your machine. The path where you checkout the repository will be denoted <checkout_dir>/models
git clone
  1. Add the directory <checkout_dir>/research/slim to the$PYTHONPATH variable. Or add a line to your .bashrc file.
export PYTHONPATH="<checkout_dir>/research/slim:$PYTHONPATH"
  1. Download the model checkpoints from the official page.


There are two example files, one for classification and one for feature extraction.

Feature Extraction

--network resnet_v1_101 
--checkpoint ./checkpoints/resnet_v1_101.ckpt 
--image_path ./images_dir/ 
--out_file ./features.h5
--num_classes 1000 
--layer_names resnet_v1_101/logits

--network resnet_v2_101 
--checkpoint ./checkpoints/resnet_v2_101.ckpt 
--image_path ./images_dir/
--out_file ./features.h5 
--layer_names resnet_v2_101/logits 
--preproc_func inception

--network inception_v4 
--checkpoint ./checkpoints/inception_v4.ckpt 
--image_path ./images_dir/
--out_file ./features.h5 
--layer_names Logits

Image Classification
--network resnet_v1_101 
--checkpoint ./checkpoints/resnet_v1_101.ckpt 
--image_path ./images_dir/
--num_classes 1000 
--logits_name resnet_v1_101/logits

Work in Progress

  1. Save image file names to HDF5 file
  2. Support for multi-threaded preprocessing