Update README.md

apple2373 · web-flow · commit 7230284827ff · 2019-01-15T14:43:16.000-05:00
diff --git a/README.md b/README.md
@@ -19,3 +19,4 @@ please see minimum-workingexample.ipynb
 
 This is just a minimum working example to show how our automatic road mask generation works, which is implemented solely by me (Satoshi Tsutsui). Because I did the work in an internship and  still do not have the company's permission to open source the code, I reproduced from scratch by myself.  For the experiments in the paper, we used other implementation with gpu enabled cupy backed, and the core part is mostly coded by Shunta Saito (my co-author). As far as I recognize, a big difference is the input resolution. While this one use 224 x 224 for input, the paper used higher resolution for superpixels algorithms so that we can get more precise superpixels. 
 
+Update: My PFN co-author made the official one available: https://github.com/pfnet-research/superpixel-align

Original file line number	Diff line number	Diff line change
`@@ -19,3 +19,4 @@ please see minimum-workingexample.ipynb`
`19`	`19`
`20`	`20`	This is just a minimum working example to show how our automatic road mask generation works, which is implemented solely by me (Satoshi Tsutsui). Because I did the work in an internship and still do not have the company's permission to open source the code, I reproduced from scratch by myself. For the experiments in the paper, we used other implementation with gpu enabled cupy backed, and the core part is mostly coded by Shunta Saito (my co-author). As far as I recognize, a big difference is the input resolution. While this one use 224 x 224 for input, the paper used higher resolution for superpixels algorithms so that we can get more precise superpixels.
`21`	`21`
	`22`	`+Update: My PFN co-author made the official one available: https://github.com/pfnet-research/superpixel-align`