def main(args=None):
# parse arguments
if args is None:
args = sys.argv[1:]
args = parse_args(args)
# make sure keras is the minimum required version
check_keras_version()
# optionally choose specific GPU
if args.gpu:
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
keras.backend.tensorflow_backend.set_session(get_session())
# make save path if it doesn't exist
if args.save_path is not None and not os.path.exists(args.save_path):
os.makedirs(args.save_path)
# create the generator
generator = create_generator(args)
# load the model
print('Loading model, this may take a second...')
# model = keras.models.load_model(args.model, custom_objects=custom_objects)
#
inputs = keras.layers.Input(shape=(None, None, 3))
resnet = keras_resnet.models.ResNet50(inputs, include_top=False, freeze_bn=True)
model = retinanet.retinanet_bbox(inputs=inputs, num_classes=65, backbone=resnet)
model.load_weights(args.model)
# print model summary
print(model.summary())
# start evaluation
average_precisions, average_recalls = evaluate(
generator,
model,
iou_threshold=args.iou_threshold,
score_threshold=args.score_threshold,
max_detections=args.max_detections,
top=args.top,
wordvector=args.wordvector,
save_path=args.save_path
)
# print evaluation
for label, average_precision in average_precisions.items():
print(generator.label_to_name(label), '{:.4f}'.format(average_precision))
seenMAP = sum(average_precisions.values()[:65]) / 65
unseenMAP = sum(average_precisions.values()[65:]) / 15
HMMAP = (2 * seenMAP * unseenMAP) / (seenMAP + unseenMAP)
seenRE = sum(average_recalls.values()[:65]) / 65
unseenRE = sum(average_recalls.values()[65:]) / 15
HMRE = (2 * seenRE * unseenRE) / (seenRE + unseenRE)
print('Seen mAP: {:.4f} Unseen mAP: {:.4f} HM mAP: {:.4f}'.format(seenMAP,unseenMAP, HMMAP))
print('Seen Rec: {:.4f} Unseen Rec: {:.4f} HM Rec: {:.4f}'.format(seenRE, unseenRE, HMRE))
def main(args=None):
# parse arguments
if args is None:
args = sys.argv[1:]
args = parse_args(args)
# make sure keras is the minimum required version
check_keras_version()
# optionally choose specific GPU
if args.gpu:
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
keras.backend.tensorflow_backend.set_session(get_session())
# make save path if it doesn't exist
if args.save_path is not None and not os.path.exists(args.save_path):
os.makedirs(args.save_path)
# create the generator
generator = create_generator(args)
# load the model
print('Loading model, this may take a second...')
# model = keras.models.load_model(args.model, custom_objects=custom_objects)
# shafin
inputs = keras.layers.Input(shape=(None, None, 3))
resnet = keras_resnet.models.ResNet50(inputs,
include_top=False,
freeze_bn=True)
model = retinanet.retinanet_bbox(inputs=inputs,
num_classes=72,
backbone=resnet)
model.load_weights(args.model)
print("args_model:{}".format(args.model))
# print model summary
print(model.summary())
# start evaluation
# average_precisions, average_recalls
all_detections_with_labels = evaluate(generator,
model,
iou_threshold=args.iou_threshold,
score_threshold=args.score_threshold,
max_detections=args.max_detections,
top=args.top,
wordvector=args.wordvector,
save_path=args.save_path)
# print(
# 'all_detections_with_labels:{}\nshape:{}'.format(all_detections_with_labels, all_detections_with_labels.shape)
# )
if not os.path.exists('detection.h5'):
with h5py.File('detection.h5') as f:
f['dets_labels'] = all_detections_with_labels
def main(args=None):
# parse arguments
if args is None:
args = sys.argv[1:]
args = parse_args(args)
# make sure keras is the minimum required version
check_keras_version()
# optionally choose specific GPU
if args.gpu:
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
keras.backend.tensorflow_backend.set_session(get_session())
# create the generators
train_generator, validation_generator = create_generators(args)
if 'resnet' in args.backbone:
from keras_retinanet.models import retinanet
from ..models.resnet_vocab_w2v import custom_objects
from keras_retinanet.models import retinanet_vocab_w2v
import keras_resnet.models
elif 'mobilenet' in args.backbone:
from ..models.mobilenet import mobilenet_retinanet as retinanet, custom_objects, download_imagenet
else:
raise NotImplementedError('Backbone \'{}\' not implemented.'.format(args.backbone))
# create the model
if args.snapshot is not None:
print('Loading model, this may take a second...')
model = keras.models.load_model(args.snapshot, custom_objects=custom_objects)
training_model = model
prediction_model = model
else:
inputs = keras.layers.Input(shape=(None, None, 3))
resnet = keras_resnet.models.ResNet50(inputs, include_top=False, freeze_bn=True)
training_model = retinanet_vocab_w2v.retinanet_bbox(inputs=inputs, num_classes=72, backbone=resnet)
# compile model
training_model.compile(
loss={
'regression' : losses.smooth_l1(),
'classification': losses.polar()
},
optimizer=keras.optimizers.adam(lr=1e-5, clipnorm=0.001)
)
# Loading a Pre-trained Model
# inputs = keras.layers.Input(shape=(None, None, 3))
# cocomodel_resnet = keras_resnet.models.ResNet50(inputs, include_top=False, freeze_bn=True)
# cocomodel65 = retinanet.retinanet_bbox(inputs=inputs, num_classes=72, backbone=cocomodel_resnet)
# cocomodel65.load_weights('coco72/resnet50_csv_11.h5')
#
# for i in range(0, cocomodel65.layers.__len__(), 1):
# wei = cocomodel65.layers[i].get_weights()
# try:
# training_model.layers[i].set_weights(wei)
# except Exception as e:
# print i
training_model.load_weights('Model/resnet50_csv_30.h5')
print('yes')
print 'Pre-trained load done'
print(training_model.summary())
# create the callbacks
callbacks = create_callbacks(
training_model,
training_model,
training_model,
validation_generator,
args,
)
# start training
training_model.fit_generator(
generator=train_generator,
steps_per_epoch=args.steps,
epochs=args.epochs,
verbose=1,
callbacks=callbacks,
)
num_seen = 65
word = np.loadtxt('MSCOCO/word_w2v.txt', dtype='float32', delimiter=',')
word_seen = word[:, :num_seen]
word_unseen = word[:, num_seen:]
wordname_lines = open('MSCOCO/cls_names_test_coco.csv').read().split("\n")
class_mapping = {}
for idx in range(int(len(wordname_lines)) - 1):
class_mapping[idx] = wordname_lines[idx].split(',')[0]
inputs = keras.layers.Input(shape=(None, None, 3))
resnet = keras_resnet.models.ResNet50(inputs,
include_top=False,
freeze_bn=True)
model = retinanet.retinanet_bbox(inputs=inputs,
num_classes=num_seen,
backbone=resnet)
model.load_weights('Model/resnet50_polar_loss.h5')
lines = open('sample_input.txt').read().split("\n")
num_rois = 100
visualise = False # False True
detect_type = 'gzsd' # gzsd or zsd or seen_detection
seen_threshold = .4
unseen_threshold = .2
for idx in range(int(len(lines)) - 1):
aline = lines[idx].split(" ")
im_id = aline[1]
filepath = aline[0]
