It was a research project that I was interested in working on. After making a little progress, I hit the wall on hardware resource limitations. Additionally, I was bombarded with new papers week by week becoming the state-of-the-art in supervised edge detection, last one being the Holisticaly-nested Edge Detection completely crushing the competitions (last year's competition, though) :D

Due to these problems, and the fact that I was working on other research projects, I stopped working on this project. However, after skimming through recent submission to CVPR 2016 and ICLR 2016, I'm seeing new datasets and a new-born interest in this problem. This had made me think twice about this project and I think I'll be resuming work on SuperEdge in the (hopefully) near future.