-
Notifications
You must be signed in to change notification settings - Fork 181
/
cfgs.py
75 lines (59 loc) · 2 KB
/
cfgs.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
71
72
# -*- coding: utf-8 -*-
from __future__ import division, print_function, absolute_import
import numpy as np
from alpharotate.utils.pretrain_zoo import PretrainModelZoo
from configs._base_.models.retinanet_r50_fpn import *
from configs._base_.datasets.dota_detection import *
from configs._base_.schedules.schedule_1x import *
# schedule
BATCH_SIZE = 1
GPU_GROUP = "0"
NUM_GPU = len(GPU_GROUP.strip().split(','))
SAVE_WEIGHTS_INTE = 27000
DECAY_STEP = np.array(DECAY_EPOCH, np.int32) * SAVE_WEIGHTS_INTE
MAX_ITERATION = SAVE_WEIGHTS_INTE * MAX_EPOCH
WARM_SETP = int(WARM_EPOCH * SAVE_WEIGHTS_INTE)
# model
# backbone
pretrain_zoo = PretrainModelZoo()
PRETRAINED_CKPT = pretrain_zoo.pretrain_weight_path(NET_NAME, ROOT_PATH)
TRAINED_CKPT = os.path.join(ROOT_PATH, 'output/trained_weights')
# loss
CENTER_LOSS_MODE = 1 # center loss in kld
CLS_WEIGHT = 1.0
REG_WEIGHT = 0.01
VERSION = 'RetinaNet_DOTA_KF_KL_1x_20220918_v2'
"""
RetinaNet-H + log(kl_center) + kfiou (exp(1-IoU)-1)
loss = (loss_1.reshape([n, 1]) + loss_2).reshape([n*n,1])
loss = sum(loss)
loss /= n
This is your evaluation result for task 1 (VOC metrics):
mAP: 0.7159506032559942
ap of each class:
plane:0.8956319958448199,
baseball-diamond:0.7853195018991056,
bridge:0.4565355846786333,
ground-track-field:0.6866889276919379,
small-vehicle:0.7542019498311174,
large-vehicle:0.7271779236220344,
ship:0.8447137519356132,
tennis-court:0.9081264105818754,
basketball-court:0.8137256266611681,
storage-tank:0.7965270049452231,
soccer-ball-field:0.5417951637064815,
roundabout:0.6304533082486643,
harbor:0.6442105194223064,
swimming-pool:0.6937147697639319,
helicopter:0.560436610007
COCO style result:
AP50: 0.7159506032559942
AP75: 0.39637966915928124
mAP: 0.4005802343000219
The submitted information is :
Description: RetinaNet_DOTA_KF_KL_1x_20220918_v2_35.1w
Username: yangxue
Institute: DetectionTeamUCAS
Emailadress: yangxue16@mails.ucas.ac.cn
TeamMembers: yangxue, yangjirui
"""