Committed instruction

This commit is contained in:
Chaoyue Wang
2017-07-19 21:27:21 -04:00
parent 43caf9d65d
commit c443a7d8f1
+6 -3
View File
@@ -11,16 +11,19 @@ Download converted weights here:
And place in directory with pspnet50_ade20k.npy
Weights from caffemodel were converted by, weight_converter.py. The usage of this file is
``bash
```bash
python weight_converter.py <path to .prototxt> <path to .caffemodel>
``
```
Running this need to compile the original PSPNet caffe code and pycaffe.
Interpolation layer is implemented in code as custom layer "Interp"
## Important
This repo is forked from [Vladkryvoruchko/PSPNet-Keras-tensorflow](https://github.com/Vladkryvoruchko/PSPNet-Keras-tensorflow) and repaired some issues.
This repo is forked from [Vladkryvoruchko/PSPNet-Keras-tensorflow](https://github.com/Vladkryvoruchko/PSPNet-Keras-tensorflow) and repaired some issues. Here is an example of current code:
![Original](test.jpg)
![Processed](test_seg.jpg)
![Alpha mixed](test_seg_blended.jpg)
## Dependencies:
1. Tensorflow