Object Detection using Transfer Learning

Custom object detection model using tensorflow object detection api - Momo Detector

Petrained Model

ssd_mobilenet_v1_coco (from tensorflow model zoo)

Dependencies:

• Labelimg - to label images
• Protobufs - to configue model and training parameters.

Directories:

• Data directory contains xml files generated by labelimg, csv files converted from xml and pbtxt file for mapping labels.
• momo-images conains momo jpeg images.
• momo_inference_graph contains trained model with checkpoints.
• test_images contains images used to test.
• utils - utility files used by ipynb files to visualize output.
• ipynb file contains sample visualization of the model.