研究文章
通过DenseNet Multi-Indices量化为左心室,GRU-Based Encoder-Decoder与关注
|
| 模块 |
Pytorch层 |
在渠道 |
从渠道 |
型芯尺寸 |
步 |
填充 |
偏见 |
|
| conv1 |
Conv2d |
1 |
16 |
(3) |
(1,1) |
(1,1) |
假 |
|
| 瓶颈(0) |
BatchNorm2d |
16 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
| Conv2d |
16 |
32 |
(1,1) |
(1,1) |
(0,0) |
假 |
| BatchNorm2d |
32 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
| Conv2d |
32 |
8 |
(3) |
(1,1) |
(1,1) |
假 |
|
| 瓶颈(1) |
BatchNorm2d |
24 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
| Conv2d |
24 |
32 |
(1,1) |
(1,1) |
(0,0) |
假 |
| BatchNorm2d |
32 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
| Conv2d |
32 |
8 |
(3) |
(1,1) |
(1,1) |
假 |
|
| 过渡 |
BatchNorm2d |
32 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
| Conv2d |
32 |
16 |
(1,1) |
(1,1) |
(0,0) |
假 |
|
| bn1 |
BatchNorm2d |
32 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
|
| 池 |
avg_pool2d |
32 |
- - - - - - |
(4,4) |
(4,4) |
(0,0) |
- - - - - - |
|
| 足球俱乐部 |
线性 |
512年 |
128年 |
- - - - - - |
- - - - - - |
- - - - - - |
- - - - - - |
|
|