PaddleSeg

PaddleSeg copied to clipboard

### 问题确认 Search before asking - [X] 我已经查询[历史issue](https://github.com/PaddlePaddle/PaddleSeg/issues)(包含open与closed)，没有类似需求。I have searched the [open and closed issues](https://github.com/PaddlePaddle/PaddleSeg/issues) and found no similar feature requests. ### 需求描述 Feature Description 1. 任务目标（请描述你正在做的项目是什么，如模型、论文、项目是什么？） 用于小目标分割 2. 需求场景（请描述你的项目中为什么需要用此功能）;...

zouxiaodong

Unable to Export or Perform Inference on PanopticSeg Models -> Mask2Former and Panoptic-DeepLab

2

### 问题确认 Search before asking - [X] 我已经查询[历史issue](https://github.com/PaddlePaddle/PaddleSeg/issues)(包括open与closed)，没有发现相似的bug。I have searched the [open and closed issues](https://github.com/PaddlePaddle/PaddleSeg/issues) and found no similar bug report. ### Bug描述 Describe the Bug **current dir:** /content/drive/MyDrive/Arslan/paddle_seg/PaddleSeg/contrib/PanopticSeg **Performing...

Arslan-Mehmood1

### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 我在Ubuntu服务器中创建了paddle2.5.1和paddleseg的虚拟环境，在终端运行代码时一切正常，但是用在pycharm中点击运行或者调试就会报这个错误，请问怎么解决呢  。ImportError: /home/xxx/anaconda3/envs/paddle/lib/python3.7/site-packages/paddle/fluid/libpaddle.so: undefined symbol: cudaGraphDebugDotPrint,...

Kerwinji

使用PP-Matting-512 模型，在T4 显卡gpu，抠图执行时间20s正常么

1

### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 使用PP-Matting-512 模型，在T4 显卡gpu，抠图执行时间20s正常么

xiaoyaogaobin

实用deeplabv3p进行训练报错，使用的是自定义数据集（已反复监检测无误）：cv2.error: OpenCV(4.5.5) :-1: error: (-5:Bad argument) in function 'resize'

1

### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 报错：cv2.error: OpenCV(4.5.5) :-1: error: (-5:Bad argument) in...

in-moonshadow

variance is not in pretrained model

1

### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 使用pplite-seg进行训练后，加载我自己训练的模型的时候报错： 2023-11-27 11:48:40 [INFO] Loading pretrained model...

jasper-cell

人像分割模型pp-humanseg-mobile分割出来的人像往往会包含一些背景内容，该怎么优化边缘呢？

1

### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 人像分割模型pp-humanseg-mobile分割出来的人像往往会包含一些背景内容，该怎么优化边缘呢？ 

xingxinghanzi

视频图像抠图功能在推理时会出现多余的size信息打印

1

### 问题确认 Search before asking - [X] 我已经查询[历史issue](https://github.com/PaddlePaddle/PaddleSeg/issues)(包括open与closed)，没有发现相似的bug。I have searched the [open and closed issues](https://github.com/PaddlePaddle/PaddleSeg/issues) and found no similar bug report. ### Bug描述 Describe the Bug 执行命令： ``` python tools/bg_replace_video.py...

15675110562

unet3+训练过程中评估，直接退出

3

### 问题确认 Search before asking - [X] 我已经查询[历史issue](https://github.com/PaddlePaddle/PaddleSeg/issues)(包括open与closed)，没有发现相似的bug。I have searched the [open and closed issues](https://github.com/PaddlePaddle/PaddleSeg/issues) and found no similar bug report. ### Bug描述 Describe the Bug but right dtype is...

KangLifei

Got low loss when training but it predicted a whole white image with the size of input image then ROI

1

我尝试用提供的optic数据测试我的代码，训练跟预测都没有问题，但是每次换成我的数据后就开始出现问题，我的图像也是jpg，标注数据也是png并且名字里面没有空格。就是训练不了，之前还没问题，突然不知道是怎么了。 I have tried the provided optic_dis_seg data, it is working fine. But when I switch to my own data this issue starts to pop out. my image is also...

SonoHistology