/tf_badminton

A project to read the in-out of badminton shuttlecock using the tensorflow object detection API

Primary LanguagePython

Tensorflow API를 이용한 프로젝트

TF_Badminton IN-OUT Lineman Project

A project to read the in-out of badminton shuttlecock using the tensorflow object detection API
Created on a Colab basis.
Use tensorflow object detection API

1. Setup

  1. drive mount

2. Setting the Work Environment

  1. import the models for object detection
  • tensorflow ssd mobilenet
  1. add tf libraries to pythonpath
  2. download pre-trained model
  • ssd_mobilenet_v2_quantized_300x300_coco_2019_01_03

3. Update model pipeline configurations

  1. model
  • Line 3: num_classes = number of classes in your dataset.
  1. train_config
  • Line 157: fine_tune_checkpoint = MODEL_DIR/training/model.ckpt
  1. train_input_reader
  • Line 162: label_map_path = MODEL_DIR/data/labelmap.pbtxt
  • Line 164: input_path = MODEL_DIR/data/train.record
  1. eval_config
  • Line 168: num_examples = number of test images
  1. eval_input_reader
  • Line 174: label_map_path = MODEL_DIR/data/labelmap.pbtxt
  • Line 178: input_path = MODEL_DIR/data/test.record

4. Train

  • if ckpt-0: 37:36[click to run] ~ 39:35[activate first step] => about 2 min to start learning (2019-12-14)
  • num_train_steps=100000

5. Tflite converting

title Download cocoapi_clone

  • helped metrics per category successfully on github
  • cocoapi_clone from some user for us
  • Without it, 'export_tflite_ssd_graph.py' doesn't work