-
-
1）PaddlePaddle版本：2.2
2）GPU：1080 ti， cuda 10.2，cudnn 7.6.5
3）系统环境：Windows 10 专业版19042.1165， python 3.9
运行nvcc -V：
C:\Users\HUAOPT>nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 200…
-
For the computation of dice scores and the loss, what do you take as the optic disc?
Is it the entire optic disc (including the cup) or is it just the "ring" (not including the cup), i.e. where the…
-
python export.py \
--config configs/quick_start/bisenet_optic_disc_512x512_1k.yml \
--model_path output/iter_1000/model.pdparams
报错提示：AttributeError: 'Assign' object has no attrib…
-
_base_: '../_base_/cityscapes.yml'
batch_size: 2
iters: 8000
model:
type: DeepLabV3P
backbone:
type: MobileNetV3_large_x1_0
output_stride: 32
h/MobileNetV3_large_x1_0.tar.gz
n…
-
环境：win10,cuda11,paddleseg2.3,paddlex2.0,bisenetv2
模型导出命令：python export.py --config configs\quick_start/bisenet_optic_disc_512x512_1k-100000.yml --model_path output\best_model\model.pdparams --input…
-
paddle版本v2
python版本3.7
标注工具labelme
原因1
使用tools/labelme2seg.py得到的label图像不能用np.asarray(Image.open(label))得到灰度图,
自己的数据用np.unique(label[:,:,0]),np.unique(label[:,:,1]),np.unique(label[:,:,2])查看得到B通…
L-lei updated
3 years ago
-
请教个关于背景颜色和伪彩图的问题：
通常背景像素默认为0，对应到paddleseg里边的伪彩图颜色应该是[128,0,0]，但我看到paddleseg用来做例子的数据集optic_disc_seg里边的伪彩图，背景的伪彩颜色为[0,0,0]也就是黑色，那么到底在paddleseg里边对背景像素是怎么认定的，[128,0,0]和[0,0,0]都是背景伪彩色吗？没有匹配正确的背景色会不会导致训练错误…
-
Error: Blocking queue is killed because the data reader raises an exception
[Hint: Expected killed_ != true, but received killed_:1 == true:1.] at (D:\1.8.1\paddle\paddle/fluid/operators/reader/block…
-
I am wondering about the synonyms on Optic nerve coloboma HP:0000588
Is "Optic disk[disc] coloboma" truly synonymous with optic nerve coloboma? If not, this may be contributing to UMLS mixing the ne…