Refactored MTCNN codebase with significant optimizations. Version 1.0.0

[ipazc]

• Completely refactored the MTCNN implementation following best coding practices.
• Optimized code by removing unnecessary transpositions, resulting in faster computation. Fixes Why flip dx1, dy1, dx2, dy2 in generateBoundingBox when y.shape[0]==1? Please help me~ #22.
• Transposed convolutional layer weights to eliminate the need for additional transpositions during preprocessing and postprocessing, improving overall efficiency.
• Converted preprocessing and postprocessing functions into matrix operations to accelerate computation. Fixes how to increase worker cpu  #14, Performance is not what it could be #110.
• Added batch processing support to enhance performance for multiple input images. Fixes Passing batch of images for faster processing #9, Predicting On Batch #71.
• Migrated network architecture to TensorFlow >= 2.12 for improved compatibility and performance. Fixes Thread bug #80, Multiple detection and tf.function retracing #82, Initialize MTCNN with ValueError #90, ('Tensor Tensor("conv2d_21/BiasAdd:0", shape=(?, ?, ?, 4), dtype=float32) is not an element of this graph.',) #91, Cannot import name 'Softmax' #93, Cudnn PoolForward launch failed exception #98, Error with tf-nightly-gpu 2.5.0.dev20201207 #104, cannot import name 'get_config' from 'tensorflow.python.eager.context' #112, AttributeError: module 'keras.utils.generic_utils' has no attribute 'to_snake_case' #114, Error in importing #115, AttributeError: module 'keras.utils.generic_utils' has no attribute 'populate_dict_with_module_objects' #116.
• Extensively documented the project with detailed explanations of thresholds and parameters. Fixes Understand min_face_size param in Mtcnn #12, error with my own picture #41, nipples being detected as eyes #52, threshold #57, How to tweak the parameters of MTCNN to improve the accuracy #99, mtcnn: self.thresholds, self.nms_thresholds,self.factor What does it mean? How do I tweak it? #122, Get wrong detected face #117.
• Added support for selecting computation backends (CPU, GPU, etc.) with the device parameter. Fixes Taking too long to process images. How to leverage GPU properly? #23.
• Added new parameters to control the result format (support for x1, y1, x2, y2 instead of x1, y1, width, height) and the ability to return tensors instead of dictionaries. Fixes Box coordinates #72.
• Configured PyLint support to ensure code quality and style adherence.
• Organized functions into specific modules (mtcnn.utils.* and mtcnn.stages.*) for better modularity.
• Created Jupyter notebooks for visualization and ablation studies of each stage, allowing detailed exploration of layers, weights, and intermediate results. Fixes Extract weights #88, How to get face embedding / encoding? #102.
• Added a comprehensive training guide for the model. Fixes How to Train the Pre-trained Model? #35, Request for sharing the code for training networks? #39.
• Updated README with information on the new version, including the complete Read the Docs documentation that describes the process, theoretical background, and usage examples. Fixes Regarding the MIT License #53, ONET or RNET is the final layer?  #73.
• Configured GitHub Actions for continuous integration and delivery (CI/CD).
• Fixed memory leak by switching to a more efficient TensorFlow method (model(tensor) instead of model.predict(tensor)). Fixes Memory Leak #87, Memory leak from mtcnn #109, Disable progress bars? #121, process is killed automatically #125, Prevent mtcnn logging #128.
• Made TensorFlow an optional dependency to prevent conflicts with user-installed versions. Fixes Tensorflow-gpu already installed , pip install tries to install tensorflow #95.
• Added comprehensive unit tests for increased reliability and coverage.