- Transfer learning
- YoloV4 configuration(Inference only)
- YoloV4 training
- YoloV4 loss function adjustments.
- Live plot losses
- Command line handler
- YoloV3 tiny
- Rasberry Pi support
-
- (new) DarkNet models loaded directly from .cfg files
- (new) YoloV4 support
- tensorflow-2.X--keras-functional-api
- cpu-gpu support
- Random weights and DarkNet weights support
- csv-xml annotation parsers.
- Anchor generator.
matplotlibvisualization of all stages.tf.datainput pipeline.pandas&numpydata handling.imgaugaugmentation pipeline(customizable).loggingcoverage.- All-in-1 custom trainer.
- Stop and resume training support.
- Fully vectorized mAP evaluation.
labelpixsupport.- Photo & video detection
Here are the packages you'll need to install before starting to use the detector:
- pandas==1.0.3
- lxml==4.5.0
- opencv_python_headless==4.2.0.34
- imagesize==1.2.0
- seaborn==0.10.0
- tensorflow==2.2.0
- tensorflow-gpu==2.2.0
- numpy==1.18.2
- matplotlib==3.2.1
- imgaug==0.4.0
- tensorflow-addons==0.10.0
- imagecorruptions==1.1.0
- Clone the repo
git clone https://github.com/emadboctorx/yolov3-keras-tf2/- Install requirements
pip install -r requirements.txtor
conda install --file requirements.txtyolov3-keras-tf2 is initially an implementation of yolov3 (you only look once)(training & inference) and YoloV4 support was added(02/06/2020) which is is a state-of-the-art, real-time object detection system that is extremely fast and accurate.There are many implementations that support tensorflow, only a few that support tensorflow v2 and as I did not find versions that suit my needs so, I decided to create this version which is very flexible and customizable. It requires the Python interpreter version 3.6, 3.7, 3.7+, is not platform specific and is MIT licensed which means you can use, copy, modify, distribute this software however you like.
- YoloV4 training is currently supported.
- Models are loaded directly from DarkNet .cfg files
- YoloV4 is currently supported(inference only, no training)
- Mish activation function(YoloV4)
This feature was introduced to replace the old hard-coded model, models are currently loaded directly from DarkNet .cfg files for convenience including YoloV4 .cfg
As models currently load from .cfg files directly, YoloV4 is supported the configuration file needs to be supplied and the model is loaded, as there are technical issues encountered with the loss function, only inference using DarkNet weights for YoloV4 is currently supported. the configuration file needs to be supplied and the model is loaded and ready for usage.
This program leverages features that were introduced in tensorflow 2.0 including:
- Eager execution: an imperative programming environment that evaluates operations immediately, without building graphs check here
tf.function: A JIT compilation decorator that speeds up some components of the program check heretf.data: API for input pipelines check here
The program detects and uses available GPUs at runtime(training/detection) if no GPUs available, the CPU will be used(slow).
Both options are available, and NOTE in case of using DarkNet yolov3 weights you must maintain the same number of COCO classes (80 classes) as transfer learning to models with different classes will be supported in future versions of this program.
There are 2 currently supported formats that the program is able to read and translate to input.
- XML VOC format which looks like the following example:
<annotation>
<folder>/path/to/image/folder</folder>
<filename>image_filename.png</filename>
<path>/path/to/image/folder/image_filename.png</path>
<size>
<width>1344</width>
<height>756</height>
<depth>3</depth>
</size>
<object>
<name>Car</name>
<bndbox>
<xmin>873.0000007680001</xmin>
<ymin>402.0000001920001</ymin>
<xmax>1315.00000128</xmax>
<ymax>697.0000000320001</ymax>
</bndbox>
</object>
<object>
<name>Car</name>
<bndbox>
<xmin>550.999999872</xmin>
<ymin>404.999999838</ymin>
<xmax>883.000000512</xmax>
<ymax>711.000000018</ymax>
</bndbox>
</object>
<object>
<name>Car</name>
<bndbox>
<xmin>8.999999903999992</xmin>
<ymin>374.999999976</ymin>
<xmax>525.99999984</xmax>
<ymax>736.000000344</ymax>
</bndbox>
</object>
<object>
<name>Traffic Lights</name>
<bndbox>
<xmin>857.999999808</xmin>
<ymin>312.99999960599996</ymin>
<xmax>903.9999991679999</xmax>
<ymax>372.99999933</ymax>
</bndbox>
</object>
<object>
<name>Traffic Lights</name>
<bndbox>
<xmin>1220.99999952</xmin>
<ymin>91.999999854</ymin>
<xmax>1317.999999456</xmax>
<ymax>249.99999985799997</ymax>
</bndbox>
</object>
<object>
<name>Traffic Lights</name>
<bndbox>
<xmin>701.999999232</xmin>
<ymin>207.00000014399998</ymin>
<xmax>753.999998976</xmax>
<ymax>275.000000184</ymax>
</bndbox>
</object>
<object>
<name>Street Sign</name>
<bndbox>
<bndbox>
<xmin>1220.99999952</xmin>
<ymin>91.999999854</ymin>
<xmax>1317.999999456</xmax>
<ymax>249.99999985799997</ymax>
</bndbox>
</object>
<object>
<name>Traffic Lights</name>
<bndbox>
<xmin>701.999999232</xmin>
<ymin>207.00000014399998</ymin>
<xmax>753.999998976</xmax>
<ymax>275.000000184</ymax>
</bndbox>
</object>
<object>
<name>Street Sign</name>
<bndbox>
<xmin>798.99999984</xmin>
<ymin>244.999999944</ymin>
<xmax>881.00000016</xmax>
<ymax>275.000000184</ymax>
</bndbox>
</annotation>- CSV with relative labels that looks like the following example:
| Image | Object Name | Object Index | bx | by | bw | bh | # |
|---|---|---|---|---|---|---|---|
| img1.png | dog | 2 | 0.438616071 | 0.51521164 | 0.079613095 | 0.123015873 | |
| img1.png | car | 1 | 0.177827381 | 0.381613757 | 0.044642857 | 0.091269841 | |
| img2.png | Street Sign | 5 | 0.674107143 | 0.44047619 | 0.040178571 | 0.084656085 |
A k-means algorithm finds the optimal sizes and generates anchors with process visualization.
Including:
- k-means visualization:
- Generated anchors:
- Precision and recall curves:
- Evaluation bar charts:
- Actual vs. detections:
- Augmentation options visualization:
Double screen visualization(before/after) image like the following example:
- Dataset pre and post augmentation visualization with bounding boxes:
You can always visualize different stages of the program using my other repo labelpix which is tool for drawing bounding boxes, but can also be used to visualize bounding boxes over images using csv files in the format mentioned here.
TFRecords a simple format for storing a sequence of binary records. Protocol buffers are a cross-platform, cross-language library for efficient serialization of structured data and are used as input pipeline to store and read data efficiently the program takes as input images and their respective annotations and builds training and validation(optional) TFRecords to be further used for all operations and TFRecords are also used in the evaluation(mid/post) training, so it's valid to say you can delete images to free space after conversion to TFRecords.
Most of the operations are using numpy and pandas for efficiency and vectorization.
Special thanks to the amazing imgaug creators, an augmentation pipeline(optional) is available and NOTE that the augmentation is conducted before the training not during the training due to technical complications to integrate tensorflow and imgaug. If you have a small dataset, augmentation is an option and it can be preconfigured before the training check Augmentor.md
Different operations are recorded using logging module.
For custom training, Trainer class accepts configurations for augmentation,
new anchor generation, new dataset(TFRecord(s)) creation, mAP evaluation
mid-training and post training. So all you have to do is place images
in Data > Photos, provide the configuration that suits you and start the training
process, all operations are managed from the same place for convenience.
For detailed instructions check Trainer.md
by default the trainer checkpoints to Models > checkpoint_name.tf at the end of each training epoch which enables the training to be resumed at any given point by loading the checkpoint which would be the most recent.
Evaluation is optional during the training every n epochs(not recommended for large datasets as it predicts every image in the dataset) and one evaluation at the end which is optional as well. Training and validation datasets can be evaluated separately and calculate mAP(mean average precision) as well as precision and recall curves for every class in the model, check Evaluator.md
You can check my other repo labelpix which is a labeling tool for drawing bounding boxes over images if you need to make custom datasets the tool can help and is supported by the detector. You can use csv files in the format mentioned here as labels and load images if you need to preview any stage of the training/augmentation/evaluation/detection.
Detections can be performed on photos or videos using Predictor class check Predictor.md
Here are the most basic steps to train using a custom dataset:
1- Copy images to Data > Photos
2- If labels are in the XML VOC format, copy label xml files to Data > Labels
3- Create classes .txt file that contains classes delimited by \n
dog
cat
car
person
boat
fan
laptop
4- Create a training instance and specify input_shape, classes_file,
image_width and image_height
trainer = Trainer(
input_shape=(416, 416, 3),
model_configuration='../Config/yolo3.cfg',
classes_file='/path/to/classes_file.txt',
image_width=1344, # The original image width
image_height=756 # The original image height
)
5- Create dataset configuration(dict) that contains the following keys:
dataset_name: TFRecord prefix(required)
and one of the following:(required)
relative_labels: path to csv file in the following format
or
from_xml:True
and
test_size: percentage of the validation split ex: 0.1(optional)augmentation:True(optional)
and if augmentation this implies the following:
-
sequences: (required) A list of augmentation sequences check Augmentor.md -
aug_workers: (optional) defaults to 32 parallel augmentations. -
aug_batch_size: (optional) this is the augmentation batch size defaults to 64 images to load at once.dataset_conf = { 'relative_labels': '/path/to/labels.csv', 'dataset_name': 'dataset_name', 'test_size': 0.2, 'sequences': preset_1, # check Config > augmentation_options.py 'augmentation': True, }
6- Create new anchor generation configuration(dict) that contains the following keys:
anchor_no: number of anchors(should be 9) and one of the following:-
relative_labels: same as dataset configuration above -
from_xml: same as dataset configuration aboveanchors_conf = { 'anchor_no': 9, 'relative_labels': '/path/to/labels.csv' }
-
7- Start the training
Note
If you're going to use DarkNet yolov3 weights, make sure the classes file contains 80 classes(COCO classes) or you'll get an error. Transfer learning to models with different number of classes will be supported in future versions of the program.
tr.train(epochs=100,
batch_size=8,
learning_rate=1e-3,
dataset_name='dataset_name',
merge_evaluation=False,
min_overlaps=0.5,
new_dataset_conf=dataset_conf, # check step 5
new_anchors_conf=anchors_conf, # check step 6
# weights='/path/to/weights' # If you're using DarkNet weights or resuming training
)
After the training completes:
- The trained model is saved in Models folder(which you can use to resume training later/predict photos or videos)
- The resulting TFRecords and their corresponding csv data are saved in Data > TFRecords
- The resulting figures and evaluation results are saved in Output folder.
Here are the most basic steps to augment images(no training, just augmentation):
If you need to augment photos and take your time to examine/visualize the results, here are the steps:
1- Copy images to Data > Photos or specify image_folder param
2- Ensure you have a csv file containing the labels in the format
mentioned here, if you have labels
in xml VOC format, you can easily convert them using Helpers > annotation_parsers.py >
parse_voc_folder() (everything is explained in the docstrings)
3- Create augmentation instance:
from Config.augmentation_options import augmentations
from Helpers.augmentor import DataAugment
aug = DataAugment(
labels_file='/path/to/labels/csv/file',
augmentation_map=augmentations)
aug.create_sequences(sequences) # check the docs
aug.augment_photo_folder()
After augmentation you'll find augmented images in the Data > Photos folder or the folder you specified(if you did specify one)
And you should find 2 csv files in the Output folder:
-
augmented_data_plus_original.csv: you can use this with labelpix to visualize results with bounding boxes -
adjusted_data_plus_original.csv
and any of the 2 csv files above can be used in the new dataset configuration in the training.
Here are the most basic steps to evaluate a trained model:
-
Create an evaluation instance:
evaluator = Evaluator( input_shape=(416, 416, 3), model_configuration='../Config/yolo3.cfg', train_tf_record='/path/to/train.tfrecord', valid_tf_record='/path/to/valid.tfrecord', classes_file='/path/to/classes.txt', anchors=anchors, # defaults to yolov3 anchors score_threshold=0.1 # defaults to 0.5 but it's okay to be lower ) -
Read actual and prediction results(that resulted from the training)
actual = pd.read_csv('../Data/TFRecords/full_data.csv') preds = pd.read_csv('../Output/full_dataset_predictions.csv') -
Calculate mAP(mean average precision):
evaluator.calculate_map( prediction_data=preds, actual_data=actual, min_overlaps=0.5, display_stats=True)
After evaluation, you'll find resulting plots and predictions in the Output folder.
Here are the most basic steps to perform detection:
-
Create a detection instance:
p = Detector( input_shape=(416, 416, 3), model_configuration='/path/to/DarkNet/yolo_version.cfg, '/path/to/classes_file.txt', score_threshold=0.5, iou_threshold=0.5, max_boxes=100, anchors=anchors # Optional if not specified, yolo default anchors are used ) -
Perform detections:
A) Photos:
photos = ['photo/path1', 'photo/path2']
p.predict_photos(photos=photos,
trained_weights='/path/to/trained/weights') # .tf or yolov3.weights(80 classes)
B) Video
p.detect_video(
'/path/to/target/vid',
'/path/to/trained/weights.tf',
)
After predictions is complete you'll find photos/video in Output > Detections
Contributions are what make the open source community such an amazing place to
learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the Project
- Create your Feature Branch (
git checkout -b feature/AmazingFeature) - Commit your Changes (
git commit -m 'Add some AmazingFeature') - Push to the Branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Give a ⭐️ if this project helped you!
Emad Boctor - emad_1989@hotmail.com
Project link: https://github.com/emadboctorx/yolov3-keras-tf2