-
Notifications
You must be signed in to change notification settings - Fork 4
/
util.py
70 lines (58 loc) · 1.93 KB
/
util.py
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
#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
@author: ZuoXiang
@contact: zx_data@126.com
@file: util.py
@time: 2020/4/28 17:26
@desc:
"""
import os
os.environ['CUDA_VISIBLE_DEVICES'] = '1'
import torch
import Resnet_101
from PIL import Image
from main import parser
from csn import ConditionalSimNet
from torchvision import transforms
from tripletnet_outfit import CS_Tripletnet
args = parser.parse_args()
def init_model(checkpoint_path):
"""
load model from checkpoint file.
:param checkpoint_path: the path of checkpoint file. (String)
:return: the model object.
"""
criterion = torch.nn.TripletMarginLoss(margin=args.margin)
model = Resnet_101.resnet34(pretrained=True, embedding_size=args.dim_embed)
csn_model = ConditionalSimNet(model, n_conditions=args.num_concepts,
embedding_size=args.dim_embed, learnedmask=args.learned, prein=args.prein)
tnet = CS_Tripletnet(csn_model, criterion)
checkpoint = torch.load(checkpoint_path)
tnet.load_state_dict(checkpoint['state_dict'])
tnet.eval()
tnet.cuda()
return tnet
cloth_match_model_women = init_model("./runs/test_20/checkpoint_16.pth.tar")
cloth_match_model_men = init_model("./runs/test_20/checkpoint_16.pth.tar")
print("Load model and data successful!")
def load_image(image_path):
"""
load image and transform image to pytorch Tensor.
:param image_path: the path of image.(String)
:return: Transformed image tensor.
"""
image = Image.open(image_path)
# RGBA to RGB
if image.mode != "RGBA":
image = image.convert("RGBA")
# image normalization
# normalize = transforms.Normalize(mean=[0.485, 0.456, 0.406],
# std=[0.229, 0.224, 0.225])
transform = transforms.Compose([
transforms.Scale(384),
transforms.CenterCrop(384),
transforms.ToTensor(),
])
image = transform(image)
return image