encoder_decoder modified to avoid bugs when running PSPNet.

getting_started.md bug fixed.

Author: johnzja

docs/getting_started.md

@@ -338,7 +338,7 @@ The final output filename will be `psp_r50_512x1024_40ki_cityscapes-{hash id}.pt
 We provide a script to convert model to [ONNX](https://github.com/onnx/onnx) format. The converted model could be visualized by tools like [Netron](https://github.com/lutzroeder/netron). Besides, we also support comparing the output results between Pytorch and ONNX model.
 
 ```shell
-python tools/pytorch2onnx.py ${CONFIG_FILE} --checkpoint ${CHECKPOINT_FILE} --output_file ${ONNX_FILE} [--shape ${INPUT_SHAPE} --verify]
+python tools/pytorch2onnx.py ${CONFIG_FILE} --checkpoint ${CHECKPOINT_FILE} --output-file ${ONNX_FILE} [--shape ${INPUT_SHAPE} --verify]
 ```
 
 **Note**: This tool is still experimental. Some customized operators are not supported for now.

mmseg/models/segmentors/encoder_decoder.py

@@ -196,7 +196,7 @@ def slide_inference(self, img, img_meta, rescale):
                 count_mat[:, :, y1:y2, x1:x2] += 1
         assert (count_mat == 0).sum() == 0
         # We want to regard count_mat as a constant while exporting to ONNX
-        count_mat = torch.from_numpy(count_mat.detach().numpy())
+        count_mat = torch.from_numpy(count_mat.cpu().detach().numpy())
         preds = preds / count_mat
         if rescale:
             preds = resize(