This is source code for Visual data recognition subject

How to run:

Type the below command for code run procedure:

python reid.py -h

You should get:

usage: reid.py [-h] [-b] [-p PATH] [-t TYPE] [-m METRIC] [-r RANK]

Process some integers.

optional arguments:
  -h, --help            show this help message and exit
  -b, --benchmark       run benchmark
  -p PATH, --path PATH  path to image
  -t TYPE, --type TYPE  feature type
  -m METRIC, --metric METRIC
                        distance metric
  -r RANK, --rank RANK  rank

1. Data preparation:

wget -O market1501 https://www.dropbox.com/s/qllazogolszz2hq/39965_62075_bundle_archive.zip?dl=0
unzip market1501

2. Install enviroment, Ubuntu and Anaconda are required:

conda env create -f environment.yml
conda activate re-id

3. Run Benchmark:

We provide several features as below:

Feature	encode name
Flatten raw image	naive
Histogram of LBP	lbp
Histogram of BGR	BGR_hist
Histogram of HSV	HSV_hist
Histogram of HS	HS_hist
Histogram of HSV split 2_0	HSV_hist_2_0
Histogram of BGR split 2_0	BGR_hist_2_0
Histogram of BGR split 2_2	BGR_hist_2_2
Histogram of HSV split 2_2	HSV_hist_2_2
Histogram of BGR split 4_0	BGR_hist_4_0
Histogram of HSV split 4_0	HSV_hist_4_0

We also provide 2 kind of Distance calculation:

Distance	encode name
Euclide Distnace	l2_distance
histogram_intersection	histogram_intersection

You can run experiment by:

python reid.py -b -t <encode name of Feature> -m <encode name of metric>

Example:

python reid.py -b -t HSV_hist_4_0 -m histogram_intersection

4. Run test:

You can run test by:

python reid.py -p <path to test image> -t <encode name of Feature> -m <encode name of metric> -r <rank>

Example:

python reid.py -p ./example/0001_c6s1_009601_00.jpg -t HSV_hist_4_0 -m histogram_intersection -r 15

The result will be store at: result.png
You should get:

Benchmark result:

Methods	Rank1	Rank5	Rank10	Rank15	Rank20	Feature extraction time(s)/ img	Feature comparation time (s)/img
Naïve image_L2	0.0035	0.0115	0.0187	0.0243	0.0285	9.7095e-06	0.9008
BGR_H00_HI	0.1722	0.2704	0.3206	0.3536	0.3773	0.00058	0.4577
BGR_H20_HI	0.2520	0.3690	0.4266	0.4649	0.4907	0.00085	0.4992
BGR_H22_HI	0.2449	0.3580	0.4094	0.4441	0.4667	0.00139	0.5971
BGR_H40_HI	0.3147	0.4323	0.4928	0.5240	0.5489	0.00141	0.5963
HSV_H00_HI	0.3450	0.4922	0.5540	0.5899	0.6166	0.00065	0.4729
HS_H00_HI	0.2687	0.4192	0.4925	0.5279	0.5596	0.00048	0.4483
HSV_H20_HI	0.4314	0.5748	0.6350	0.6612	0.6873	0.00093	0.51549
HSV_H22_HI	0.4035	0.5498	0.6054	0.6395	0.6633	0.00157	0.63320
HSV_H40_HI	0.4685	0.6184	0.6852	0.7119	0.7301	0.00179	0.72152
LBP_H00_HI	0.0742	0.1401	0.1802	0.2051	0.2244	0.00130	0.43157
LPB_H00_L2	0.0492	0.1045	0.1419	0.1677	0.1840	0.00129	0.16462

thanhson28 / cv_subject Goto Github PK

cv_subject's Introduction

How to run:

1. Data preparation:

2. Install enviroment, Ubuntu and Anaconda are required:

3. Run Benchmark:

4. Run test:

Benchmark result:

cv_subject's People

Contributors

Stargazers

Watchers

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent