This is the Tensorflow (Version 0.11) implementation of CVPR-17 paper "Deep Visual-Semantic Quantization for Efficient Image Retrieval". The descriptions of files in this directory are listed below:
net.py
: contains the main implementation (network structure, loss function, optimization procedure and etc.) of the proposed approachdvsq
.net_val.py
: contains the implementation ofdvsq
for evaluation.util.py
: contains the implementation ofDataset
,MAP
andProcessBar
.train_script.py
: gives an example to show how to traindvsq
model.validation_script.py
: gives an example to show how to evaluate the trained quantization model.run_dvsq.sh
: gives an example to show the full procedure of training and evaluating the proposed approachdvsq
.
In data/nuswide_81/train.txt
, we give an example to show how to prepare image training data. In data/nuswide_81/test.txt
and data/nuswide_81/database.txt
, the list of testing and database images could be processed during predicting procedure. In data/nuswide_81/nuswide_wordvec.txt
, we have already prepared the word vectors of the labels extracted by Word2Vec model pretrained on Google News Dataset.
The bvlc_reference_caffenet
is used as the pre-trained model. If the NUS_WIDE dataset and pre-trained caffemodel is prepared, the example can be run with the following command:
"./run_dvsq.sh"
@inproceedings{conf/cvpr/CaoL0L17,
author = {Yue Cao and
Mingsheng Long and
Jianmin Wang and
Shichen Liu},
title = {Deep Visual-Semantic Quantization for Efficient Image Retrieval},
booktitle = {2017 {IEEE} Conference on Computer Vision and Pattern Recognition,
{CVPR} 2017, Honolulu, Hawaii, USA, July 21-26, 2017}
}