Skip to content

thtang/Sparse-Coding-for-Face-Image-Retrieval

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sparse-Coding-for-Face-Image-Retrieval

Data

LFW DATA Labeled Faces in the Wild (LFW) is a widely used benchmark for face verification. It contains 13233 images of 5749 different people. You can find more information on the data set at http://vis-www.cs.umass.edu/lfw/.

Usage

Download the dataset.

cd ./src
python3 face_image_retrieval.py [LFW pickle data] [LFW attribute txt]

Dependency

Python3 numpy joblib scipy pandas matplotlib spams

Baseline result

Experiment report

Compute L2 distance of local binary pattern (LBP) and rank.

Query 1st    2nd 3rd 4th 5th

Sparse coding approach

Compute Cosine similarity of Spare coding with identity information and rank.

Query 1st    2nd 3rd 4th 5th
  • mAP of vanilla sparse coding

  • mAP of sparse coding with identity information

Reference

[1] Scalable Face Image Retrieval using Attribute-Enhanced Sparse Codewords, Chen et al., IEEE Trans. Multimedia 2013

[2] Semi-supervised face image retrieval using sparse coding with identity constraint, Chen et al., ACM MM 2011

[3] SPArse Modeling Software