Improved README

Author: liebharc

README.md

@@ -1,14 +1,14 @@
 # homr
 
-homr is a Optical Music Recognition (OMR) software which takes camera pictures of sheet music and turns them into machine readable [MusicXML](https://www.w3.org/2021/06/musicxml40/). MusicXML can then futher processed e.g. with tools such as [musescore](https://musescore.com/).
+homr is an Optical Music Recognition (OMR) software which takes camera pictures of sheet music and turns them into machine readable [MusicXML](https://www.w3.org/2021/06/musicxml40/). MusicXML can then futher processed e.g. with tools such as [musescore](https://musescore.com/).
 
 ## Technical Details
 
-homr uses the segmentation as described in [oemer](https://github.com/BreezeWhite/oemer) to find possible staffs, clefs, bar lines and notes on an image. It combines the detects staffs, clefs and bars to find the position of the staffs in the picture.
+homr uses the segmentation as described in [oemer](https://github.com/BreezeWhite/oemer) to find possible staff lines, clefs, bar lines and note heads on an image. It combines the staff lines, clefs and bars to find the position of the staffs in the picture.
 
 In an next step if passes each staff image into the transformer (based on [Polyphonic-TrOMR](https://github.com/NetEase/Polyphonic-TrOMR/tree/master/tromr/model)) to get the symbols on the staff.
 
-Note information is double checked with the note information obtained from the segmentation model.
+Pitch information is double checked with the note head information obtained from the segmentation model.
 
 The results are then converted into MusicXML and written to disk.