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