forked from yangxue0827/RotationDetection
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcfgs_res50_dota1.5_kf_v3.py
105 lines (84 loc) · 2.78 KB
/
cfgs_res50_dota1.5_kf_v3.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
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
# -*- 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 = 32000
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)
# dataset
DATASET_NAME = 'DOTA1.5'
CLASS_NUM = 16
# 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
CLS_WEIGHT = 1.0
REG_WEIGHT = 0.01
VERSION = 'RetinaNet_DOTA1.5_KF_1x_20210904'
"""
RetinaNet-H + kfiou (exp(1-IoU)-1)
loss = (loss_1.reshape([n, 1]) + loss_2).reshape([n*n,1])
loss = sum(loss)
loss /= n
FLOPs: 485785312; Trainable params: 33051321
This is your evaluation result for task 1:
mAP: 0.621280142030018
ap of each class:
plane:0.7961673552273434,
baseball-diamond:0.7919297962108256,
bridge:0.4266645800680319,
ground-track-field:0.6725785853434287,
small-vehicle:0.49442312033891705,
large-vehicle:0.617605960472413,
ship:0.7589145831913362,
tennis-court:0.8948022695795292,
basketball-court:0.746940541069233,
storage-tank:0.618297423618113,
soccer-ball-field:0.5107664025542281,
roundabout:0.6868455436184394,
harbor:0.6161102848881593,
swimming-pool:0.6437627881067691,
helicopter:0.5731867467072281,
container-crane:0.09148629148629149
The submitted information is :
Description: RetinaNet_DOTA1.5_KF_1x_20210904_41.6w
Username: SJTU-Det
Institute: SJTU
Emailadress: [email protected]
TeamMembers: yangxue
This is your evaluation result for task 1:
mAP: 0.62712466195339
ap of each class:
plane:0.7996320679842436,
baseball-diamond:0.7931046992887474,
bridge:0.4311646095372172,
ground-track-field:0.681974776888092,
small-vehicle:0.500308973139615,
large-vehicle:0.6235146979358908,
ship:0.7652515264086934,
tennis-court:0.9081722581874745,
basketball-court:0.7604111462221048,
storage-tank:0.622222185541879,
soccer-ball-field:0.5132890348503273,
roundabout:0.6924873883332214,
harbor:0.6229200343597264,
swimming-pool:0.6513720079418596,
helicopter:0.5766821592638252,
container-crane:0.09148702537132289
The submitted information is :
Description: RetinaNet_DOTA1.5_KF_1x_20210904_41.6w_cpunms
Username: SJTU-Det
Institute: SJTU
Emailadress: [email protected]
TeamMembers: yangxue
"""