File size: 10,573 Bytes
fd4b932
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
# train.py

import argparse
import os
os.sys.path += ['expman']
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import math
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflowjs as tfjs
from tensorflow.keras import backend as K
from tensorflow.keras.models import load_model
from tensorflow.keras.callbacks import ModelCheckpoint, LearningRateScheduler, CSVLogger
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report, roc_curve, auc, precision_recall_curve, average_precision_score
from adabelief_tf import AdaBeliefOptimizer
from tqdm.keras import TqdmCallback
from tqdm import tqdm
from functools import partial

from dataloader import get_loader, load_datasets, validate_data_files
from models.unet import build_model
from utils import visualize
from expman import Experiment
import evaluate

def boundary_loss(y_true, y_pred):
    """Additional loss focusing on boundaries"""
    y_true = tf.cast(y_true, tf.float32)
    y_pred = tf.cast(y_pred, tf.float32)
    # Compute gradients
    dy_true, dx_true = tf.image.image_gradients(y_true)
    dy_pred, dx_pred = tf.image.image_gradients(y_pred)

    # Compute boundary loss
    loss = tf.reduce_mean(tf.abs(dy_pred - dy_true) + tf.abs(dx_pred - dx_true))
    return loss * 0.5

def enhanced_binary_crossentropy(y_true, y_pred):
    """Combine standard BCE with boundary loss"""
    y_true = tf.cast(y_true, tf.float32)
    y_pred = tf.cast(y_pred, tf.float32)
    bce = tf.keras.losses.binary_crossentropy(y_true, y_pred)
    boundary = boundary_loss(y_true, y_pred)
    return bce + boundary

def cosine_decay_with_warmup(epoch, total_epochs, warmup_epochs=5, initial_lr=0.001):
    if epoch < warmup_epochs:
        # Linear warmup
        return initial_lr * (epoch + 1) / warmup_epochs
    # Cosine decay after warmup
    progress = (epoch - warmup_epochs) / (total_epochs - warmup_epochs)
    return initial_lr * (1 + math.cos(math.pi * progress)) / 2

def main(args):
    try:
        # Verify data directories exist
        for data_dir in args.data:
            if not os.path.exists(data_dir):
                raise FileNotFoundError(f"Data directory not found: {data_dir}")
        
        exp = Experiment(args, ignore=('epochs', 'resume'))
        print(exp)

        np.random.seed(args.seed)
        tf.random.set_seed(args.seed)

        data = load_datasets(args.data)
        if len(data) == 0:
            raise ValueError("No valid data found after loading datasets")

        # Validate all files exist
        validate_data_files(data)

        # TRAIN/VAL/TEST SPLIT
        if args.split == 'subjects':  # by SUBJECTS
            val_subjects = (6, 9, 11, 13, 16, 28, 30, 48, 49)
            test_subjects = (3, 4, 19, 38, 45, 46, 51, 52)
            train_data = data[~data['sub'].isin(val_subjects + test_subjects)]
            val_data = data[data['sub'].isin(val_subjects)]
            test_data = data[data['sub'].isin(test_subjects)]

        elif args.split == 'random':  # 70-20-10 %
            train_data, valtest_data = train_test_split(data, test_size=.3, shuffle=True)
            val_data, test_data = train_test_split(valtest_data, test_size=.33)

        lengths = map(len, (data, train_data, val_data, test_data))
        print("Total: {} - Train / Val / Test: {} / {} / {}".format(*lengths))

        x_shape = (args.resolution, args.resolution, 1)
        y_shape = (args.resolution, args.resolution, 1)

        train_gen, _ = get_loader(train_data, batch_size=args.batch_size, shuffle=True, augment=True, x_shape=x_shape)
        val_gen, val_categories = get_loader(val_data, batch_size=args.batch_size, x_shape=x_shape)

        log = exp.path_to('log.csv')

        # weights_only checkpoints
        best_weights_path = exp.path_to('best_weights.weights.h5')
        best_mask_weights_path = exp.path_to('best_weights_mask.weights.h5')

        # whole model checkpoints
        best_ckpt_path = exp.path_to('best_model.keras')
        last_ckpt_path = exp.path_to('last_model.keras')

        if args.resume and os.path.exists(last_ckpt_path):
            custom_objects = {
                'iou_coef': evaluate.iou_coef,
                'dice_coef': evaluate.dice_coef,
                'enhanced_binary_crossentropy': enhanced_binary_crossentropy,
                'boundary_loss': boundary_loss
            }
            model = tf.keras.models.load_model(last_ckpt_path, custom_objects=custom_objects)
            optimizer = model.optimizer
            initial_epoch = len(pd.read_csv(log)) if os.path.exists(log) else 0
        else:
            config = vars(args)
            model = build_model(x_shape, y_shape, config)

            # Use Adam optimizer
            optimizer = tf.keras.optimizers.Adam(
                learning_rate=float(args.lr),
                beta_1=0.9,
                beta_2=0.999,
                epsilon=1e-7
            )
            initial_epoch = 0

        model.compile(
            optimizer=optimizer,
            loss={
                'mask': enhanced_binary_crossentropy,
                'tags': 'binary_crossentropy'
            },
            metrics={
                'mask': [evaluate.iou_coef, evaluate.dice_coef],
                'tags': 'binary_accuracy'
            }
        )

        model_stopped_file = exp.path_to('early_stopped.txt')
        need_training = not os.path.exists(model_stopped_file) and initial_epoch < args.epochs

        if need_training:
            lr_schedule = partial(cosine_decay_with_warmup,
                    total_epochs=args.epochs,
                    warmup_epochs=5,
                    initial_lr=args.lr)

            best_checkpointer = ModelCheckpoint(
                best_weights_path,
                monitor='val_loss',
                save_best_only=True,
                save_weights_only=True,
                mode='min'
            )

            best_mask_checkpointer = ModelCheckpoint(
                best_mask_weights_path,
                monitor='val_mask_dice_coef',
                mode='max',
                save_best_only=True,
                save_weights_only=True
            )

            last_checkpointer = ModelCheckpoint(
                last_ckpt_path,
                save_best_only=False,
                save_weights_only=False
            )

            logger = CSVLogger(log, append=args.resume)
            progress = TqdmCallback(verbose=1, initial=initial_epoch, dynamic_ncols=True)

            early_stop = tf.keras.callbacks.EarlyStopping(
                monitor='val_mask_dice_coef',
                mode='max',
                patience=100,
                restore_best_weights=True
            )

            lr_scheduler = LearningRateScheduler(lr_schedule)

            callbacks = [
                best_checkpointer,
                best_mask_checkpointer,
                last_checkpointer,
                logger,
                progress,
                early_stop,
                lr_scheduler
            ]

            try:
                model.fit(
                    train_gen,
                    epochs=args.epochs,
                    callbacks=callbacks,
                    initial_epoch=initial_epoch,
                    steps_per_epoch=len(train_gen),
                    validation_data=val_gen,
                    validation_steps=len(val_gen),
                    verbose=False
                )
            except Exception as e:
                print(f"Training failed: {str(e)}")
                raise

            if model.stop_training:
                open(model_stopped_file, 'w').close()

            # Save the model in .keras format
            best_ckpt_path = exp.path_to('best_model.keras')
            tf.keras.models.save_model(model, best_ckpt_path, include_optimizer=False)

            # Only evaluate if training was successful
            evaluate.evaluate(exp, force=need_training)

            # save best snapshot in SavedModel format
            model.load_weights(best_mask_weights_path)
            best_savedmodel_path = exp.path_to('best_savedmodel')
            model.save(best_savedmodel_path, save_traces=True)

            # export to tfjs (Layers model)
            tfjs_model_dir = exp.path_to('tfjs')
            tfjs.converters.save_keras_model(model, tfjs_model_dir)
        else:
            print("No training needed, model already exists and training completed.")
            # Optionally evaluate existing model
            evaluate.evaluate(exp, force=False)

    except Exception as e:
        print(f"Error in main: {str(e)}")
        raise

if __name__ == '__main__':
    default_data = ['data/NN_human_mouse_eyes']

    parser = argparse.ArgumentParser(description='MEye Training Script')
    # data params
    parser.add_argument('-d', '--data', nargs='+', default=default_data, help='Data directory (may be multiple)')
    parser.add_argument('--split', default='random', choices=('random', 'subjects'), help='How to split data')
    parser.add_argument('-r', '--resolution', type=int, default=128, help='Input image resolution')

    # model params
    parser.add_argument('--num-stages', type=int, default=5, help='number of down-up sample stages')
    parser.add_argument('--num-conv', type=int, default=1, help='number of convolutions per stage')
    parser.add_argument('--num-filters', type=int, default=16, help='number of conv filter at first stage')
    parser.add_argument('--grow-factor', type=float, default=1.5,
                        help='# filters at stage i = num-filters * grow-factor ** i')
    parser.add_argument('--up-activation', default='relu', choices=('relu', 'lrelu'),
                        help='activation in upsample stages')
    parser.add_argument('--conv-type', default='conv', choices=('conv', 'bn-conv', 'sep-conv', 'sep-bn-conv'),
                        help='convolution type')
    parser.add_argument('--use-aspp', default=False, action='store_true', help='Use Atrous Spatial Pyramid Pooling')

    # train params
    parser.add_argument('--lr', type=float, default=0.001, help='learning rate')
    parser.add_argument('-b', '--batch-size', type=int, default=32, help='Batch size')
    parser.add_argument('-e', '--epochs', type=int, default=1500, help='Number of training epochs')
    parser.add_argument('-s', '--seed', type=int, default=23, help='Random seed')
    parser.add_argument('--resume', default=False, action='store_true', help='Resume training')

    args = parser.parse_args()
    main(args)