Intro
This repository contains a CBIR (content-based image retrieval) system
Extract query image's feature, and retrieve similar ones from image database
Image src
Part1: Feature Extraction
In this system, I implement several popular image features:
- color-based
- texture-based
- shape-based
- deep methods
all features are modulized
Feature Fusion
Some features are not robust enough, and turn to feature fusion
Dimension Reduction
The curse of dimensionality told that vectors in high dimension will sometimes lose distance property
Part2: Evaluation
CBIR system retrieves images based on feature similarity
Robustness of system is evaluated by MMAP (mean MAP), the evaluation formulas is refer to here
- image AP : average of precision at each hit
- depth=K means the system will return top-K images
- a correct image in top-K is called a hit
- AP = (hit1.precision + hit2.precision + ... + hitH.precision) / H
- class1 MAP = (class1.img1.AP + class1.img2.AP + ... + class1.imgM.AP) / M
- MMAP = (class1.MAP + class2.MAP + ... + classN.MAP) / N
Implementation of evaluation can found at evaluate.py
My database contains 25 classes, each class with 20 images, 500 images in total, depth=K will return top-K images from database
| Method | color | daisy | edge | gabor | HOG | vgg19 | resnet152 |
|---|---|---|---|---|---|---|---|
| Mean MAP (depth=10) | 0.614 | 0.468 | 0.301 | 0.346 | 0.450 | 0.914 | 0.944 |
Part3: Image Retrieval (return top 5 of each method)
Let me show some results of the system
query1 - women dress
query 
color 
daisy 
edge 
gabor 
HOG 
VGG19 
Resnet152 
query2 - orange
query 
color 
daisy 
edge 
gabor 
HOG 
VGG19 
Resnet152 
query3 - NBA jersey
query 
color 
daisy 
edge 
gabor 
HOG 
VGG19 
Resnet152 
query4 - snack
query 
color 
daisy 
edge 
gabor 
HOG 
VGG19 
Resnet152 
Part4: Usage of Repository
If you are interesting with the results, and want to try your own images,
Please refer to USAGE.md
The details are written inside.
Author
Po-Chih Huang / @pochih