This is an implementation of Faster RCNN on BCCD image set. The data set contains of roughly around 400 microscopic images of blood along with xml files containing the annotations and information about bounding boxes. We have to identify and localize, in an image, whether a cell is RBC, WBC or Platelet. I used Tensorflow object detection api for this. You can find the dataset here https://github.com/Shenggan/BCCD_Dataset