when I run demo .py, occurs a bug 'Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)'

[wenjun2022]

MAC os Pycharm 源码： from mmseg.apis import inference_segmentor, init_segmentor import mmcv import torch

print(torch.version.cuda) config_file = r'/Users/wenjun/pythonProject/mmsegmentation-master/configs/setr/setr_vit-large_pup_8x1_768x768_80k_cityscapes.py' checkpoint_file = '/Users/wenjun/pythonProject/mmsegmentation-master/checkpoints/setr_pup_vit-large_8x1_768x768_80k_cityscapes_20211122_155115-f6f37b8f.pth'

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

从一个 config 配置文件和 checkpoint 文件里创建分割模型

model = init_segmentor(config_file, checkpoint_file, device=device)

测试一张样例图片并得到结果

img = r'/Users/wenjun/pythonProject/mmsegmentation-master/demo/demo.png' # 或者 img = mmcv.imread(img), 这将只加载图像一次． result = inference_segmentor(model, img)

在新的窗口里可视化结果

model.show_result(img, result, show=True)

或者保存图片文件的可视化结果

您可以改变 segmentation map 的不透明度(opacity)，在(0, 1]之间。

model.show_result(img, result, out_file='result.jpg', opacity=0.5)

Run： /Users/wenjun/miniconda3/envs/open-mmlaba/bin/python /Users/wenjun/pythonProject/mmsegmentation-master/demo/test.py None

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

[MengzhangLI]

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV) always happened by different packages conflict. Could you run any code in MMSegmentation? If not, you can create another new conda environment follow here.

[MengzhangLI]

Another possible cause (which I encountered today) is that you're trying to read/write a file which is open. In this case, simply closing the file and rerunning the script solved the issue. from here. Do you open demo.png when implementing demo.py?