StanCallewaert StanCallewaert
StanCallewaert
/ convert_for_edge_tpu_use.sh
Last active
April 3, 2019 12:13
Convert our current model to a model for the Edge TPU
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| python object_detection/export_tflite_ssd_graph.py \ | |
| --pipeline_config_path="pipeline.config" \ | |
| --trained_checkpoint_prefix="model.ckpt-500" \ | |
| --output_directory="output" \ | |
| --add_postprocessing_op=true | |
| bazel run -c opt tensorflow/contrib/lite/toco:toco -- \ | |
| --input_file=output/tflite_graph.pb \ | |
| --output_file=output/detect.tflite \ | |
| --input_shapes=1,300,300,3 \ |
StanCallewaert
/ package_object_detection.sh
Last active
April 3, 2019 12:13
Package the Object Detection API, pycocotools, and TF Slim
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| bash object_detection/dataset_tools/create_pycocotools_package.sh /tmp/pycocotools | |
| python setup.py sdist | |
| (cd slim && python setup.py sdist) |
StanCallewaert
/ train_model.sh
Last active
April 3, 2019 12:10
Train the Tensorflow detection model
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| gcloud ml-engine jobs submit training training_job \ | |
| --job-dir=gs://edge-tpu-demo/train \ | |
| --packages dist/object_detection-0.1.tar.gz,slim/dist/slim-0.1.tar.gz,/tmp/pycocotools/pycocotools-2.0.tar.gz | |
| --module-name object_detection.model_tpu_main \ | |
| --runtime-version 1.9 \ | |
| --scale-tier BASIC_TPU \ | |
| --region europe-west4 \ | |
| -- \ | |
| --model_dir=gs://edge-tpu-demo/train | |
| --tpu_zone europe-west4 \ |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| item { | |
| id: 1 | |
| name: 'cutter' | |
| } | |
| item { | |
| id: 2 | |
| name: 'plier' | |
| } | |
| item { | |
| id: 3 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| model { | |
| ssd { | |
| num_classes: 3 | |
| ... | |
| train_config { | |
| batch_size: 24 | |
| fine_tune_checkpoint: "gs://edge-tpu-demo/model.ckpt" | |
| freeze_variables: [ | |
| 'FeatureExtractor/MobilenetV2/Conv', | |
| 'FeatureExtractor/MobilenetV2/expanded_conv_1', |
StanCallewaert
/ data_to_tfrecords.sh
Last active
April 3, 2019 12:14
Convert our data to the TFRecords file format
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| git clone git@github.com:datitran/raccoon_dataset.git | |
| # Load your images and annotations in the correct folder | |
| rm raccoon_dataset/annotations/* raccoon_dataset/images/* | |
| mv image*.xml raccoon_dataset/annotations/ | |
| mv image*.jpg raccoon_dataset/images/ | |
| # Convert your XML files to a CSV file | |
| cd raccoon_dataset | |
| python3 xml_to_csv.py |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| git clone git@github.com:tzutalin/labelImg.git | |
| cd labelImg | |
| brew install qt | |
| brew install libxml2 | |
| make qt5py3 | |
| python3 labelImg.py |
StanCallewaert
/ extract_images_from_video.sh
Last active
September 20, 2019 11:10
Extract images from video
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # First install ffmpeg | |
| # Then set the following 2 variables | |
| FILENAME=your_file.mp4 | |
| NUMBER_OF_IMAGES=50 | |
| VIDEO_LENGTH_IN_SECONDS=`ffprobe -i $FILENAME -show_entries format=duration -v quiet -of csv="p=0"` | |
| ffmpeg -i $FILENAME -vf fps=$NUMBER_OF_SECONDS/$NUMBER_OF_IMAGES image%04d.jpg |