-
Notifications
You must be signed in to change notification settings - Fork 181
/
Copy pathcfgs_res50_dota2.0_dcl_v5.py
77 lines (63 loc) · 2 KB
/
cfgs_res50_dota2.0_dcl_v5.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
# -*- coding: utf-8 -*-
from __future__ import division, print_function, absolute_import
import numpy as np
from configs._base_.models.retinanet_r50_fpn import *
from configs._base_.datasets.dota_detection import *
from configs._base_.schedules.schedule_1x import *
from alpharotate.utils.pretrain_zoo import PretrainModelZoo
# schedule
BATCH_SIZE = 1
GPU_GROUP = "0,1,2"
NUM_GPU = len(GPU_GROUP.strip().split(','))
SAVE_WEIGHTS_INTE = 40000 * 2
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 = 'DOTA2.0'
CLASS_NUM = 18
# 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')
# bbox head
ANGLE_RANGE = 180
# loss
CLS_WEIGHT = 1.0
REG_WEIGHT = 1.0
ANGLE_WEIGHT = 0.5
# DCL
OMEGA = 180 / 256.
ANGLE_MODE = 0 # {0: BCL, 1: GCL}
VERSION = 'RetinaNet_DOTA2.0_DCL_B_2x_20210430'
"""
FLOPs: 877875705; Trainable params: 33486966
This is your evaluation result for task 1:
mAP: 0.4545546886264481
ap of each class:
plane:0.7576838360163581,
baseball-diamond:0.48071214438681525,
bridge:0.367934070781644,
ground-track-field:0.5815236628223699,
small-vehicle:0.34513208572635395,
large-vehicle:0.36598807625753177,
ship:0.46939020389816366,
tennis-court:0.7545404991770102,
basketball-court:0.5731304185966594,
storage-tank:0.5011136945493068,
soccer-ball-field:0.4053102096300879,
roundabout:0.49786355787385084,
harbor:0.35938351489137554,
swimming-pool:0.5031132619574917,
helicopter:0.5421417282441151,
container-crane:0.12772487037593397,
airport:0.45636926716170634,
helipad:0.09292929292929293
The submitted information is :
Description: RetinaNet_DOTA2.0_DCL_B_2x_20210430_104w
Username: sjtu-deter
Institute: SJTU
Emailadress: [email protected]
TeamMembers: yangxue
"""