merged

Author: Chrps

docker/Dockerfile

@@ -28,10 +28,6 @@ RUN ["/bin/bash", "-c", "mim install mmengine"]
 RUN ["/bin/bash", "-c", "mim install mmcv==${MMCV}"]
 
 # Install MMSegmentation
-RUN echo test
-RUN echo test2
-RUN echo test3
-RUN echo test5
 RUN git clone -b main https://github.com/Chrps/mmsegmentation.git /mmsegmentation
 
 WORKDIR /mmsegmentation

docker_run.sh

@@ -1,2 +1,2 @@
 # TODO: update to make paths more general
-docker run --gpus all --shm-size=8g -it -v /media/chrps/wd_external/ambolt/PotentialProjects/Mos/mos_pascal_voc_final:/mmsegmentation/data -v /media/chrps/wd_external/ambolt/PotentialProjects/Mos/mmsegmentation/work_dirs:/mmsegmentation/work_dirs mmsegmentation
+docker run --gpus all --shm-size=8g -it -v /media/chrps/wd_external/ambolt/PotentialProjects/Mos/mos_pascal_voc_final:/mmsegmentation/data -v /media/chrps/wd_external/ambolt/PotentialProjects/Mos/mmsegmentation/work_dirs:/mmsegmentation/work_dirs mmsegmentation

inference.py

@@ -0,0 +1,23 @@
+from mmseg.apis import inference_model, init_model, show_result_pyplot
+import mmcv
+
+config_file = '/mmsegmentation/work_dirs/pascal_mos/20240402_103935/vis_data/config.py'
+checkpoint_file = '/mmsegmentation/work_dirs/pascal_mos/iter_300.pth'
+
+# build the model from a config file and a checkpoint file
+model = init_model(config_file, checkpoint_file, device='cuda:0')
+
+# test a single image and show the results
+img = '/mmsegmentation/data/JPEGImages/frame_0.jpg'  # or img = mmcv.imread(img), which will only load it once
+img = mmcv.imread(img)
+result = inference_model(model, img)
+# visualize the results in a new window
+show_result_pyplot(model, img, result, show=True)
+# or save the visualization results to image files
+# you can change the opacity of the painted segmentation map in (0, 1].
+show_result_pyplot(model, img, result, show=True, out_file='result.jpg', opacity=0.5)
+# test a video and show the results
+#video = mmcv.VideoReader('video.mp4')
+#for frame in video:
+#   result = inference_model(model, frame)
+#   show_result_pyplot(model, frame, result, wait_time=1)