Develop and train a custom object detection model using SSD MobileNet V2 FPNLite 320x320 with Tensorflow Object Detection API. Training with 2000 training steps to classify and locate 20 different classes and achieved a total loss of 0.2589 on the training set and 0.7395 on the validation set. The loss on the validation set includes 0.3866 classification loss, 0.2071 localization loss and 0.1458 regularization loss.
(Training graph)
Inference result :
* (inference_result_example-1.jpg)
* (inference_result_example-2.jpg)
Citation :
@misc{tensorflowmodelgarden2020,
author = {Hongkun Yu, Chen Chen, Xianzhi Du, Yeqing Li, Abdullah Rashwan, Le Hou, Pengchong Jin, Fan Yang,
Frederick Liu, Jaeyoun Kim, and Jing Li},
title = {{TensorFlow Model Garden}},
howpublished = {\url{https://github.com/tensorflow/models}},
year = {2020}
}