/
cfgs_res50_dota2.0_ridet_8p_v1.py
71 lines (59 loc) · 1.91 KB
/
cfgs_res50_dota2.0_ridet_8p_v1.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
# -*- 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(','))
LR = 1e-3
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')
# loss
CLS_WEIGHT = 1.0
REG_WEIGHT = 0.01
VERSION = 'RetinaNet_DOTA2.0_RIDet_2x_20210509'
"""
RIDet-8p
FLOPs: 870905839; Trainable params: 33293346
This is your evaluation result for task 1:
mAP: 0.45350021903066295
ap of each class:
plane:0.7640386782607965,
baseball-diamond:0.45941949819838707,
bridge:0.3822849941827053,
ground-track-field:0.5882485982911329,
small-vehicle:0.38182163088601057,
large-vehicle:0.35913778649186495,
ship:0.4927984243614078,
tennis-court:0.7627193407094273,
basketball-court:0.541613446379996,
storage-tank:0.521356661130177,
soccer-ball-field:0.41027757525094277,
roundabout:0.5003811125017407,
harbor:0.41894648024188597,
swimming-pool:0.5081730389824448,
helicopter:0.4385247995863651,
container-crane:0.058974774482795866,
airport:0.5203210333539394,
helipad:0.05396606925991192
The submitted information is :
Description: RetinaNet_DOTA2.0_RIDet_2x_20210509_104w
Username: yangxue
Institute: UCAS
Emailadress: yangxue16@mails.ucas.ac.cn
TeamMembers: yangxue
"""