Skip to content

Dataset of the most critically endangered species of all leopards: "Amur Leopard", that can be used to train your own amur-leopard detector using TensorFlow's Object Detection API.

License

Notifications You must be signed in to change notification settings

monilshah98/amur_leopard_dectector_and_dataset

Repository files navigation

amur_leopard_detector_and_dataset

My love and affection for amur-leopards (sad to say now the most critically endangered species of all leopards) since childhood lead me to creating this project. This is the dataset of amur leopard, that i collected to train my own amur leopard detector using TensorFlow's Object Detection API. The total dataset consists of 200 images that i collected from google and labelled it using Label Img. I used 160 images for training and 40 images for validation. I used faster_rcnn_inception_v2_coco model for training purpose.

Repository Structure:
+ CSV's: contains the label files (csv)
+ annotations: contains the xml files in PASCAL VOC format
+ test-images: contains the testing image data in jpg format
+ train-images: contains the training image data in jpg format
+ training: contains the pipeline configuration file and labelmap file
- a few scripts: generate_tfrecord.py is used to generate the input files
for the TesorFlow API and xml_to_csv.py is used to convert the xml files into one csv 
- a few jupyter notebooks: draw outline is used to plot some of the data and 
Amur_Leopard_Detection_Live_Tutorial is tutorial to test your model live using OpenCV.
+ train.record, test.record: are the input files for the TF object detection API

Copyright

See LICENSE for details. Copyright (c) 2019 Monil Shah

About

Dataset of the most critically endangered species of all leopards: "Amur Leopard", that can be used to train your own amur-leopard detector using TensorFlow's Object Detection API.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published