/CBIR-CNN

A CBIR system with CNN model

Primary LanguagePythonMIT LicenseMIT

Content Based Image Retrieval System with Convolutional Neural Networks

It is the code for a CBIR system with CNN model. We provide a demo of a web application based on Flask, which realizes the online CBIR system for receiving the uploaded images and showing the most similar top 15 pictures.

Requirements

  • Python 3.6
  • Keras 2.0.5
  • TensorFlow (1.8.0)
  • cuda & cudnn (optional, only using the feature extraction)

You can install all python requirements with:

pip install -r requirements.txt

Dataset

Caltech-256 (http://www.vision.caltech.edu/Image_Datasets/Caltech256/)

Usage

Feature extraction and build index

python index.py

Query test on local

python query.py

Calcuate the mAP

python compute_mAP.py

Run web-server

python server.py

Demo (Please wait serval seconds for downloading)

image