RocketGit

hahattpro / crowd_counting_framework (public) (License: Unspecified) (since 2020-12-23) (hash sha1)

backupfrom https://github.com/ttpro1995/crowd_counting_framework

Clone URLs: https://rocketgit.com/user/hahattpro/crowd_counting_framework ssh://rocketgit@ssh.rocketgit.com/user/hahattpro/crowd_counting_framework git://git.rocketgit.com/user/hahattpro/crowd_counting_framework

List of commits:

Subject	Hash	Author	Date (UTC)
fix padding with p	86c2fa07822d956a34b3b37e14da485a4249f01b	Thai Thien	2019-10-06 02:52:58
pacnn perspective loss	fb673e38a5f24ae9004fe2b7b93c88991e0c2304	Thai Thien	2019-10-06 01:38:28
data_flow shanghaitech_pacnn_with_perspective seem working	91d350a06f358e03223966297d124daee94123d0	Thai Thien	2019-10-06 01:31:11
multiscale loss and final loss only mode	c65dd0e74ad28503821e5c8651a3b47b4a0c7c64	Thai Thien	2019-10-05 15:58:19
wip : perspective map	eac63f2671dc5b064753acc4f40bf0f9f216ad2a	Thai Thien	2019-10-04 16:26:56
shell script	f2106e700b6f6174d4dd276f25ec6f3d9ff239bb	thient	2019-10-04 07:42:51
WIP	42c7c8e1d772fbbda61a4bdf9e329f74e1efb600	tthien	2019-10-03 17:52:47
add readme	580cf43d1edddd67b1f6a2c57fdd5cee3dba925c	Thai Thien	2019-10-02 17:44:49
update script, debug	ddb68b95389be1c1d398118677dd227a8bb2b70b	Thai Thien	2019-10-02 15:52:31
add d (output density map) to loss function)	a0c71bf4bf2ab7393d60b06a84db8dfbbfb1a6c2	tthien	2019-09-30 16:32:39
fix the args, add save interval for model, so we don't save them all	9fdf9daa2ac4bd12b7b62521d81e520db0debd01	tthien	2019-09-30 16:30:00
meow	1ad19a22a310992e27a26471feeb37375124d075	tthien	2019-09-29 18:25:43
fix pacnn perspective map	453ece3ccb818889ba895bfc4285f7905d33cba5	Thai Thien	2019-09-25 17:20:33
apex not work so well	da8c0dd57297f972201f31d57e66897177922f48	Thai Thien	2019-09-24 17:25:59
fix data loader pacnn so it will scale up with correct number of people	11d55b50d764511f2491291f0208fee0905dec49	Thai Thien	2019-09-24 15:40:56
add comet ml	a9d4b89ce594f5e241168ccafdcdf0f150ea0ebb	Thai Thien	2019-09-23 17:07:58
fix pacnn avg schema	c2140a96886195782e5689c24aeeb4fe7a2db7ad	Thai Thien	2019-09-22 17:35:01
debug number not divisible by 8	a568fd7f294a8bd31b3db78437b4b6b51b5b41b9	Thai Thien	2019-09-22 04:36:06
pacnn	967074890d14ab0eefc277801860270a468e8f9f	Thai Thien	2019-09-22 03:54:48
wip: pacnn	2192d7c7b449fecf3868877d9cfbc09bb6f7ae98	Thai Thien	2019-09-22 03:44:56

Commit 86c2fa07822d956a34b3b37e14da485a4249f01b - fix padding with p
Author: Thai Thien
Author date (UTC): 2019-10-06 02:52
Committer name: Thai Thien
Committer date (UTC): 2019-10-06 02:52
Parent(s): fb673e38a5f24ae9004fe2b7b93c88991e0c2304
Signer:
Signing key:
Signing status: N
Tree: f25ae2d50bff301267fbffb8375598d616cad8d6

File	Lines added	Lines deleted
args_util.py	4	4
data_flow.py	1	1
main_pacnn.py	18	6
train_script/train_pacnn_shanghaitechA.sh	23	3

File args_util.py changed (mode: 100644) (index fa52179..3fa2c42)
...	...	def real_args_parse():
74	74	parser.add_argument('--epochs', action="store", default=1, type=int)	parser.add_argument('--epochs', action="store", default=1, type=int)
75	75
76	76	# pacnn setting only	# pacnn setting only
77		parser.add_argument('--PACNN_PERSPECTIVE_AWARE_MODEL', action="store_true", default=False)
78		parser.add_argument('--PACNN_MUTILPLE_SCALE_LOSS', action="store", default=True,
79		help="True: compare each of density map/perspective map scale with gt for loss."
80		" False: only compare final density map and final density perspective map")
	77		parser.add_argument('--PACNN_PERSPECTIVE_AWARE_MODEL', action="store", default=0, type=int)
	78		parser.add_argument('--PACNN_MUTILPLE_SCALE_LOSS', action="store", default=1, type=int,
	79		help="1: compare each of density map/perspective map scale with gt for loss."
	80		"0: only compare final density map and final density perspective map")
81	81
82	82	# args.original_lr = 1e-7	# args.original_lr = 1e-7
83	83	# args.lr = 1e-7	# args.lr = 1e-7

File data_flow.py changed (mode: 100644) (index a7b2270..6faebf8)
...	...	def load_data_shanghaitech_pacnn_with_perspective(img_path, train=True):
118	118	img = Image.open(img_path).convert('RGB')	img = Image.open(img_path).convert('RGB')
119	119	gt_file = h5py.File(gt_path, 'r')	gt_file = h5py.File(gt_path, 'r')
120	120	target = np.asarray(gt_file['density'])	target = np.asarray(gt_file['density'])
121		perspective = np.array(h5py.File(p_path, "r")['pmap'])
	121		perspective = np.array(h5py.File(p_path, "r")['pmap']).astype(np.float32)
122	122	perspective = np.rot90(perspective, k=3)	perspective = np.rot90(perspective, k=3)
123	123	if train:	if train:
124	124	crop_size = (int(img.size[0] / 2), int(img.size[1] / 2))	crop_size = (int(img.size[0] / 2), int(img.size[1] / 2))

File main_pacnn.py changed (mode: 100644) (index 4eee4ac..823774f)
...	...	if __name__ == "__main__":
24	24
25	25	# Add the following code anywhere in your machine learning file	# Add the following code anywhere in your machine learning file
26	26	experiment = Experiment(api_key="S3mM1eMq6NumMxk2QJAXASkUM",	experiment = Experiment(api_key="S3mM1eMq6NumMxk2QJAXASkUM",
27		project_name="pacnn-dev2", workspace="ttpro1995")
	27		project_name="pacnn-dev2", workspace="ttpro1995", disabled=True)
28	28
29	29	args = real_args_parse()	args = real_args_parse()
	30		device = "cpu"
30	31	print(device)	print(device)
31	32	print(args)	print(args)
32	33
33	34
	35
34	36	MODEL_SAVE_NAME = args.task_id	MODEL_SAVE_NAME = args.task_id
35	37	MODEL_SAVE_INTERVAL = 5	MODEL_SAVE_INTERVAL = 5
36	38	DATA_PATH = args.input	DATA_PATH = args.input

...	...	if __name__ == "__main__":
50	52	experiment.log_parameter("lr", args.lr)	experiment.log_parameter("lr", args.lr)
51	53
52	54	# create list	# create list
53		if DATASET_NAME is "shanghaitech":
	55		if "shanghaitech" in DATASET_NAME:
54	56	TRAIN_PATH = os.path.join(DATA_PATH, "train_data")	TRAIN_PATH = os.path.join(DATA_PATH, "train_data")
55	57	TEST_PATH = os.path.join(DATA_PATH, "test_data")	TEST_PATH = os.path.join(DATA_PATH, "test_data")
56	58	train_list, val_list = get_train_val_list(TRAIN_PATH)	train_list, val_list = get_train_val_list(TRAIN_PATH)
57	59	test_list = create_training_image_list(TEST_PATH)	test_list = create_training_image_list(TEST_PATH)
58		elif DATASET_NAME is "ucf_cc_50":
	60		elif "ucf_cc_50" in DATASET_NAME:
59	61	train_list, val_list = get_train_val_list(DATA_PATH, test_size=0.2)	train_list, val_list = get_train_val_list(DATA_PATH, test_size=0.2)
60	62	test_list = None	test_list = None
61	63

...	...	if __name__ == "__main__":
70	72	]),	]),
71	73	train=True,	train=True,
72	74	batch_size=1,	batch_size=1,
73		num_workers=4, dataset_name="shanghaitech_pacnn"),
	75		num_workers=4, dataset_name=DATASET_NAME),
74	76	batch_size=1, num_workers=4)	batch_size=1, num_workers=4)
75	77
76	78	val_loader_pacnn = torch.utils.data.DataLoader(	val_loader_pacnn = torch.utils.data.DataLoader(

...	...	if __name__ == "__main__":
149	151
150	152	if PACNN_PERSPECTIVE_AWARE_MODEL:	if PACNN_PERSPECTIVE_AWARE_MODEL:
151	153	# TODO: loss for perspective map here	# TODO: loss for perspective map here
152		loss_p = criterion_mse(p, perspective_p) + criterion_ssim(p, perspective_p)
	154		pad_p_0 = perspective_p.size()[2] - p.size()[2]
	155		pad_p_1 = perspective_p.size()[3] - p.size()[3]
	156		p_pad = F.pad(p, (0, pad_p_1, 0, pad_p_0), mode='replicate')
	157
	158		loss_p = criterion_mse(p_pad, perspective_p) + criterion_ssim(p_pad, perspective_p)
	159
153	160	loss += loss_p	loss += loss_p
154	161	if PACNN_MUTILPLE_SCALE_LOSS:	if PACNN_MUTILPLE_SCALE_LOSS:
155		loss_p_s = criterion_mse(p_s, perspective_s) + criterion_ssim(p_s, perspective_s)
	162		pad_s_0 = perspective_s.size()[2] - p_s.size()[2]
	163		pad_s_1 = perspective_s.size()[3] - p_s.size()[3]
	164		p_s_pad = F.pad(perspective_s, (0, pad_s_1, 0, pad_s_0),
	165		mode='replicate')
	166
	167		loss_p_s = criterion_mse(p_s_pad, perspective_s) + criterion_ssim(p_s_pad, perspective_s)
156	168	loss += loss_p_s	loss += loss_p_s
157	169
158	170	# what is this, loss_d count 2 ?	# what is this, loss_d count 2 ?

File train_script/train_pacnn_shanghaitechA.sh changed (mode: 100644) (index f3944dd..a405a96)
48	48	#--task_id train_state1_attemp7	#--task_id train_state1_attemp7
49	49
50	50	#### no loss for d1, d2, d3 but only count d_final	#### no loss for d1, d2, d3 but only count d_final
	51		#python main_pacnn.py \
	52		#--input data/ShanghaiTech/part_A \
	53		#--load_model saved_model/train_state1_attemp7_180_checkpoint.pth.tar \
	54		#--epochs 300 \
	55		#--lr 1e-9 \
	56		#--PACNN_MUTILPLE_SCALE_LOSS False \
	57		#--task_id train_state1_attemp8_finalloss
	58
	59		####################
	60
51	61	python main_pacnn.py \	python main_pacnn.py \
52	62	--input data/ShanghaiTech/part_A \	--input data/ShanghaiTech/part_A \
53	63	--load_model saved_model/train_state1_attemp7_180_checkpoint.pth.tar \	--load_model saved_model/train_state1_attemp7_180_checkpoint.pth.tar \
54		--epochs 300 \
	64		--epochs 500 \
55	65	--lr 1e-9 \	--lr 1e-9 \
56		--PACNN_MUTILPLE_SCALE_LOSS False \
57		--task_id train_state1_attemp8_finalloss
	66		--PACNN_PERSPECTIVE_AWARE_MODEL 1 \
	67		--PACNN_MUTILPLE_SCALE_LOSS 1 \
	68		--task_id train_state2_attemp1
	69
	70
	71		#--input data/ShanghaiTech/part_A \
	72		#--load_model saved_model/train_state1_attemp7_180_checkpoint.pth.tar
	73		#--epochs 500
	74		#--lr 1e-9
	75		#--PACNN_PERSPECTIVE_AWARE_MODEL 1
	76		#--PACNN_MUTILPLE_SCALE_LOSS 1
	77		#--task_id dev

Hints:
Before first commit, do not forget to setup your git environment:

git config --global user.name "your_name_here"
git config --global user.email "your@email_here"

Clone this repository using HTTP(S):

git clone https://rocketgit.com/user/hahattpro/crowd_counting_framework

Clone this repository using ssh (do not forget to upload a key first):

git clone ssh://rocketgit@ssh.rocketgit.com/user/hahattpro/crowd_counting_framework

Clone this repository using git:

git clone git://git.rocketgit.com/user/hahattpro/crowd_counting_framework

You are allowed to anonymously push to this repository.
This means that your pushed commits will automatically be transformed into a merge request:

... clone the repository ...
... make some changes and some commits ...
git push origin main