sys.path.append('./resources/libraries') import os import tensorflow as tf from tensorflow.keras.optimizers.legacy import Adam from tensorflow.keras.applications import MobileNetV2 from tensorflow.keras.layers import BatchNormalization, Conv2D, Softmax, Reshape from tensorflow.keras.models import Model from ei_tensorflow.constrained_object_detection import models, dataset, metrics, util from ei_tensorflow.velo import train_keras_model_with_velo from ei_shared.pretrained_weights import get_or_download_pretrained_weights import ei_tensorflow.training WEIGHTS_PREFIX = os.environ.get('WEIGHTS_PREFIX', os.getcwd()) def build_model(input_shape: tuple, weights: str, alpha: float, num_classes: int) -> tf.keras.Model: """ Construct a constrained object detection model. Args: input_shape: Passed to MobileNet construction. weights: Weights for initialization of MobileNet where None implies random initialization. alpha: MobileNet alpha value. num_classes: Number of classes, i.e. final dimension size, in output. Returns: Uncompiled keras model. Model takes (B, H, W, C) input and returns (B, H//8, W//8, num_classes) logits. """ #! First create full mobile_net_V2 from (HW, HW, C) input #! to (HW/8, HW/8, C) output mobile_net_v2 = MobileNetV2(input_shape=input_shape, weights=weights, alpha=alpha, include_top=True) #! Default batch norm is configured for huge networks, let's speed it up for layer in mobile_net_v2.layers: if type(layer) == BatchNormalization: layer.momentum = 0.9 #! Cut MobileNet where it hits 1/8th input resolution; i.e. (HW/8, HW/8, C) cut_point = mobile_net_v2.get_layer('block_6_expand_relu') #! Now attach a small additional head on the MobileNet model = Conv2D(filters=32, kernel_size=1, strides=1, activation='relu', name='head')(cut_point.output) logits = Conv2D(filters=num_classes, kernel_size=1, strides=1, activation=None, name='logits')(model) return Model(inputs=mobile_net_v2.input, outputs=logits) def train(num_classes: int, learning_rate: float, num_epochs: int, alpha: float, object_weight: float, train_dataset: tf.data.Dataset, validation_dataset: tf.data.Dataset, best_model_path: str, input_shape: tuple, batch_size: int, use_velo: bool = False, ensure_determinism: bool = False) -> tf.keras.Model: """ Construct and train a constrained object detection model. Args: num_classes: Number of classes in datasets. This does not include implied background class introduced by segmentation map dataset conversion. learning_rate: Learning rate for Adam. num_epochs: Number of epochs passed to model.fit alpha: Alpha used to construct MobileNet. Pretrained weights will be used if there is a matching set. object_weight: The weighting to give the object in the loss function where background has an implied weight of 1.0. train_dataset: Training dataset of (x, (bbox, one_hot_y)) validation_dataset: Validation dataset of (x, (bbox, one_hot_y)) best_model_path: location to save best model path. note: weights will be restored from this path based on best val_f1 score. input_shape: The shape of the model's input batch_size: Training batch size ensure_determinism: If true, functions that may be non- deterministic are disabled (e.g. autotuning prefetch). This should be true in test environments. Returns: Trained keras model. Constructs a new constrained object detection model with num_classes+1 outputs (denoting the classes with an implied background class of 0). Both training and validation datasets are adapted from (x, (bbox, one_hot_y)) to (x, segmentation_map). Model is trained with a custom weighted cross entropy function. """ nonlocal callbacks num_classes_with_background = num_classes + 1 width, height, input_num_channels = input_shape if width != height: raise Exception(f"Only square inputs are supported; not {input_shape}") #! Use pretrained weights, if we have them for configured allowed_combinations = [{'num_channels': 1, 'alpha': 0.1}, {'num_channels': 1, 'alpha': 0.35}, {'num_channels': 3, 'alpha': 0.1}, {'num_channels': 3, 'alpha': 0.35}] weights = get_or_download_pretrained_weights(WEIGHTS_PREFIX, input_num_channels, alpha, allowed_combinations) model = build_model( input_shape=input_shape, weights=weights, alpha=alpha, num_classes=num_classes_with_background ) #! Derive output size from model model_output_shape = model.layers[-1].output.shape _batch, width, height, num_classes = model_output_shape if width != height: raise Exception(f"Only square outputs are supported; not {model_output_shape}") output_width_height = width #! Build weighted cross entropy loss specific to this model size weighted_xent = models.construct_weighted_xent_fn(model.output.shape, object_weight) prefetch_policy = 1 if ensure_determinism else tf.data.experimental.AUTOTUNE #! Transform bounding box labels into segmentation maps def as_segmentation(ds, shuffle): ds = ds.map(dataset.bbox_to_segmentation(output_width_height, num_classes_with_background)) if not ensure_determinism and shuffle: ds = ds.shuffle(buffer_size=batch_size*4) ds = ds.batch(batch_size, drop_remainder=False).prefetch(prefetch_policy) return ds train_segmentation_dataset = as_segmentation(train_dataset, True) validation_segmentation_dataset = as_segmentation(validation_dataset, False) validation_dataset_for_callback = (validation_dataset .batch(batch_size, drop_remainder=False) .prefetch(prefetch_policy)) #! Initialise bias of final classifier based on training data prior. util.set_classifier_biases_from_dataset( model, train_segmentation_dataset) if not use_velo: model.compile(loss=weighted_xent, optimizer=Adam(learning_rate=learning_rate)) #! Create callback that will do centroid scoring on end of epoch against #! validation data. Include a callback to show % progress in slow cases. callbacks = callbacks if callbacks else [] callbacks.append(metrics.CentroidScoring(validation_dataset_for_callback, output_width_height, num_classes_with_background)) callbacks.append(metrics.PrintPercentageTrained(num_epochs)) #! Include a callback for model checkpointing based on the best validation f1. callbacks.append( tf.keras.callbacks.ModelCheckpoint(best_model_path, monitor='val_f1', save_best_only=True, mode='max', save_weights_only=True, verbose=0)) if use_velo: from tensorflow.python.framework.errors_impl import ResourceExhaustedError try: train_keras_model_with_velo( model, train_segmentation_dataset, validation_segmentation_dataset, loss_fn=weighted_xent, num_epochs=num_epochs, callbacks=callbacks ) except ResourceExhaustedError as e: print(str(e)) raise Exception( "ResourceExhaustedError caught during train_keras_model_with_velo." " Though VeLO encourages a large batch size, the current" f" size of {batch_size} may be too large. Please try a lower" " value. For further assistance please contact support" " at https://forum.edgeimpulse.com/") else: model.fit(train_segmentation_dataset, validation_data=validation_segmentation_dataset, epochs=num_epochs, callbacks=callbacks, verbose=0) #! Restore best weights. model.load_weights(best_model_path) #! Add explicit softmax layer before export. softmax_layer = Softmax()(model.layers[-1].output) model = Model(model.input, softmax_layer) return model EPOCHS = args.epochs or 100 LEARNING_RATE = args.learning_rate or 0.02 BATCH_SIZE = args.batch_size or 32 model = train(num_classes=classes, learning_rate=LEARNING_RATE, num_epochs=EPOCHS, alpha=0.35, object_weight=100, train_dataset=train_dataset, validation_dataset=validation_dataset, best_model_path=BEST_MODEL_PATH, input_shape=MODEL_INPUT_SHAPE, batch_size=BATCH_SIZE, use_velo=False, ensure_determinism=ensure_determinism) disable_per_channel_quantization = False