search

IPS-LMU / EMU-webApp

The EMU-webApp is an online and offline web application for labeling, visualizing and correcting speech and derived speech data.
http://ips-lmu.github.io/EMU-webApp/
MIT License
51 stars 14 forks source link

You cant transfer your zoom from the standard to hierarchy view. #301

Open FredrikKarlssonSpeech opened 3 years ago

FredrikKarlssonSpeech commented 3 years ago

If you mark a portion of the signal and zoom into that portion, you see only the transcription that belongs to that portion of the signal. This selection is then not transferred to the hierarchy view, which I was expecting at least. This makes working in the hierarchy view more cumbersome than it needs to I think.

samgregory commented 2 years ago

Hi @FredrikKarlssonSpeech, I agree, that the hierarchy view isn't as intuitive as it could be. It turns out that implementing your suggestion was much harder than it first appeared. Do you have the means to build EMU-webApp from a git source? I'd like someone to test out what I've done and give some feedback before I make a pull request for this change: https://github.com/samgregory/EMU-webApp/tree/hierarchy-zoom

Basically, the hierarchy view doesn't match up with the labeler view because the time information isn't used. Instead each of the item labels on a level are equally spaced (unless they are collapsed). Therefore I had to make a way to calculate which items were in view and how this related to the total number on that level. I've built in a small adjustment (of 10%) because things didn't line up perfectly.

I've yet to implement the rotated transformation but it will only require a few more lines of code. Another approach that might work would be to collapse every label node that isn't in labeler view but that might confuse people.