Skip to content

Latest commit

 

History

History
31 lines (21 loc) · 1.04 KB

README.md

File metadata and controls

31 lines (21 loc) · 1.04 KB

dog_breed_classification_and_retrival

A project aimed to classify dog breeds and search for the most matched images in the database

The detail experiments and usage can be found in my website: https://bshoterj.github.io/2019/01/20/dog-retrival/

or visit my group member's github: https://github.com/INFINITSY/dog_test

Requirements

  • Python3
  • PyTorch
  • keras
  • dataset: Stanford dog dataset (18 classes) class

Demo

Golden retrieve(金毛) t1

Siberian husky(哈士奇) t2

Dobeman(杜宾) t3

Model

  • object detection: YOLOv3
  • image retrival: ResNet50 + Euclid Distance model