train_model_gait.py

import argparse
import sys
import tensorflow as tf

from model.model_gait import ModelGait
from configuration import config
from util.utils import str2bool

if __name__ == '__main__':

    parser = argparse.ArgumentParser(
        description="Arguments for train classification model")

    parser.add_argument(
        '-colab_path',
        '--colab_path',
        type=str,
        default='',
        help='colab path to data'
    )
    parser.add_argument(
        '-summary_model',
        '--summary_model',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-init_lr',
        '--init_lr',
        type=float,
        default=0.001
    )
    parser.add_argument(
        '-log_train',
        '--log_train',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-stride',
        '--stride',
        default=1,
        type=int
    )
    parser.add_argument(
        '-fc',
        '--fc',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-train',
        '--train',
        type=str2bool,
        default=True
    )
    parser.add_argument(
        '-epochs',
        '--epochs',
        default=100,
        type=int
    )
    parser.add_argument(
        '-filter_num_user',
        '--filter_num_user',
        default=None,
        type=int
    )
    parser.add_argument(
        '-model',
        '--model',
        type=str,
        default='our',
        choices=['our', 'paper']
    )
    parser.add_argument(
        '-batch_size',
        '--batch_size',
        type=int,
        default=128
    )
    parser.add_argument(
        '-flatten',
        '--flatten',
        default=False,
        type=str2bool
    )
    parser.add_argument(
        '-method',
        '--method',
        type=str,
        default='cycle_based',
        choices=['cycle_based', 'window_based', 'window_based_svm']
    )
    parser.add_argument(
        '-window_len',
        '--window_len',
        type=int
    )
    parser.add_argument(
        '-overlap',
        '--overlap',
        type=int
    )
    parser.add_argument(
        '-split',
        '--split',
        type=str,
        choices=['standard', 'paper']      
    )
    parser.add_argument(
        '-denoise',
        '--denoise',
        type=str2bool
    )
    parser.add_argument(
        '-autocorr',
        '--autocorr',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-augment_data',
        '--augment_data',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-plot_split',
        '--plot_split',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-gyroscope',
        '--gyroscope',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-train_svm',
        '--train_svm',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-only_magnitude',
        '--only_magnitude',
        type=str2bool,
        default=False
    )
    parser.add_argument(
        '-methods_aug',
        '--methods_aug',
        type=str,
        help='methods to apply in data augmentation',
        required=False)
        
    args = parser.parse_args()

    # GPU settings
    gpus = tf.config.list_physical_devices("GPU")
    if gpus:
        for gpu in gpus:
            tf.config.experimental.set_memory_growth(gpu, True)

    model = ModelGait(config, args.colab_path)
    model.load_data(filter_num_user=args.filter_num_user,method=args.method, window_len=args.window_len, overlap=args.overlap, denoise=args.denoise, autocorr=args.autocorr, gyroscope=args.gyroscope)
    model.split_train_test(method=args.method, split=args.split, plot_split=args.plot_split)
    if args.train_svm:
        model.train_svm(only_magnitude=args.only_magnitude)
    else:
        if args.augment_data:
            model.augment_train_data(methods=args.methods_aug)
        model.normalize_data()
        model.create_tf_dataset(batch_size=args.batch_size)
        model.build_model(stride=args.stride, fc=args.fc, flatten=args.flatten,
                        summary=args.summary_model, name=args.model)
        if args.train:
            model.loss_metric(init_lr=args.init_lr)
            model.train_model(log=args.log_train, epochs=args.epochs)
            model.test_model()