One issue: the auto transcript often breaks a full sentence into many small fragments. That makes shadowing less natural.
Could you either (1) let users manually merge/split segments (adjust break points), or (2) improve the segmentation algorithm?
One issue: the auto transcript often breaks a full sentence into many small fragments. That makes shadowing less natural.
Could you either (1) let users manually merge/split segments (adjust break points), or (2) improve the segmentation algorithm?