Overview / Usage

My team is trying to build a deep learning model to detect leopards in China and classificate them. The goal of the project is to detect the leopards in the images and distinguish each leopards. The semi-supervised learning model is suitable for the project. I use the Faster-RCNN to achive the goal. After working on the problem for serveral days, I notice the Mask-RCNN and I think it can increase my accuracy rate for detection. And the photos I post was the demo images of the Mask-RCNN model. My tools for developing is Intel Optimize Caffe. And I am using the Intel dev Cloud to train my datasets in VOC2012. To increase the accuracy, I will train another model for Mask-RCNN.
I hope this project can be optimized to help nonprofit organization in China protect rare animals including leopards and some other spacies.