fakhrul / dds_drone_ws

dds_drone_ws

1. Combine and copy all image to C:/GitHub/dds_drone_ws/workspace/training_dds/images/

• ensure the image file is match with xml file
• ensure there is no corrupted image - run script data_cleaning.py

cd c:\github\dds_drone_ws\script\preprocessing python data_cleaning.py

2. partition dataset for training and test

python partition_dataset.py -x -i [PATH_TO_IMAGES_FOLDER] -r 0.1

For example

run at ./scripts/preprocessing/ python partition_dataset.py -x -i /content/images -r 0.2 -o /content/

3. generate the tfrecord file

Create train data:

python generate_tfrecord.py -x [PATH_TO_IMAGES_FOLDER]/train -l [PATH_TO_ANNOTATIONS_FOLDER]/label_map.pbtxt -o [PATH_TO_ANNOTATIONS_FOLDER]/train.record

Create test data:

python generate_tfrecord.py -x [PATH_TO_IMAGES_FOLDER]/test -l [PATH_TO_ANNOTATIONS_FOLDER]/label_map.pbtxt -o [PATH_TO_ANNOTATIONS_FOLDER]/test.record

For example

python generate_tfrecord.py -x C:/GitHub/dds_drone_ws/workspace/training_dds/images/train -l C:/GitHub/dds_drone_ws/workspace/training_dds/annotations/label_map.pbtxt -o C:/GitHub/dds_drone_ws/workspace/training_dds/annotations/train.record python generate_tfrecord.py -x C:/GitHub/dds_drone_ws/workspace/training_dds/images/test -l C:/GitHub/dds_drone_ws/workspace/training_dds/annotations/label_map.pbtxt -o C:/GitHub/dds_drone_ws/workspace/training_dds/annotations/test.record

4. Run the training

• update pipeline config with number of category
• delete previous model at C:\GitHub\dds_drone_ws\workspace\training_dds\models\my_ssd_resnet50_v1_fpn\
• only left the pipeline config

python model_main_tf2.py --model_dir=models/my_ssd_resnet50_v1_fpn --pipeline_config_path=models/my_ssd_resnet50_v1_fpn/pipeline.config

5. To view the progress Open new cmd enviroment tf38, at directory trainingdds, run

tensorboard --logdir=models/my_ssd_resnet50_v1_fpn

• tips Following what people have said online, it seems that it is advisable to allow you model to reach a TotalLoss of at least 2 (ideally 1 and lower) if you want to achieve “fair” detection results. Obviously, lower TotalLoss is better, however very low TotalLoss should be avoided, as the model may end up overfitting the dataset, meaning that it will perform poorly when applied to images outside the dataset.

6. export the inference python .\exporter_main_v2.py --input_type image_tensor --pipeline_config_path .\models\my_ssd_resnet50_v1_fpn\pipeline.config --trained_checkpoint_dir .\models\my_ssd_resnet50_v1_fpn\ --output_directory .\exported-models

7. evaluate using video python evaluate_video.py