-
Notifications
You must be signed in to change notification settings - Fork 0
/
安装流程.txt
47 lines (42 loc) · 1.4 KB
/
安装流程.txt
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
conda create -n sscls python=3.8
conda install pytorch==1.10.1 torchvision==0.11.2 torchaudio==0.10.1 cudatoolkit=11.3 -c pytorch -c conda-forge
pip install openmim
mim install mmcv-full==1.5.0
mim install mmdet
mim install mmpose
pip install -r requirements.txt
pip install -e .
更改了pyskl.models.head.simple_head.py中SimpleHead中self.pool相关
太极拳源数据taichi_mod.npy
chmod 777 ***.sh
太极拳热力图数据:
https://drive.google.com/file/d/1w4trelnmqZe5F0D_aCpia-FqJ3kDMHWD/view?usp=sharing
bash order/gen_tc.sh
ncrc热力图数据:
https://ieee-dataport.org/competitions/nurse-care-activity-recognition-challenge
python project_utils/ncrc_npy.py
python project_utils/gen_ncrc_set.py
预训练模型: bash order/ntu120.sh
1. 5part_kp
2. 5part_lb
3. 17part_kp
4. 10part_kplb
太极拳的实验:
exp0: 搜寻最优的init_lr
bash order/exp0.sh
exp1: 数据预处理的影响
bash order/exp1.sh
exp2: 骨架分割策略的有效性
bash order/exp2.sh
exp3: 预训练策略的有效性
bash order/exp3.sh
exp4: 多模态late fusion的结果以及所有配置条件的结果
bash order/exp4.sh
exp5: 搜寻init_lr和预训练的作用
bash order/exp5.sh
exp6: 多配置环境下稳定结果
bash order/exp6.sh
exp7: 在ncrc数据集上的识别结果
bash order/exp7.sh
exp8: finetune条件下,多配置识别结果
bash order/exp8.sh