gohyojun15 / FINE-GRAINED-MULTI-CLASS-OBJECT-COUNTING

FINE-GRAINED MULTI-CLASS OBJECT COUNTING icip accepted

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

FINE-GRAINED-MULTI-CLASS-OBJECT-COUNTING

This repository contains KR-GRUIDAE dataset in FINE-GRAINED-MULTI-CLASS-OBJECT-COUNTING (ICIP 2021) [paper]

KR-GRUIDAE Dataset

Images of KR-GRUIDAE dataset are captured by ecologists and labeled for five classes: Red-Crowned Crane(RCC), Red-Crowned Crane Juvenile(RCCJ), White-Naped Crane(WNC), White-Naped Crane juvenile(WNCJ) and Anser albifrons (AA).

Downloads

Dataset is now available at : [LINK]

Format

  1. img (folder): This folder contains images.
  2. gt (folder): This folder contains annotation files (.mat)

we verfied the following code for loading annotation files:

from scipy.io import loadmat

mat = loadmat(MAT_FILE_PATH)

mat["RCChead"] # contains the head points of RCC
mat["RCCbody"] # contains the body points of RCC

mat["RCCJhead"] # contains the head points of RCCJ
mat["RCCJbody"] # contains the body points of RCCJ

mat["WNChead"] # contains the head points of WNC
mat["WNCbody"] # contains the body points of WNC

mat["WNCJhead"] # contains the head points of WNCJ
mat["WNCJbody"] # contains the body points of WNCJ

mat["AAhead"] # contains the head points of AA
mat["AAbody"] # contains the body points of AA

About

FINE-GRAINED MULTI-CLASS OBJECT COUNTING icip accepted