Internet Technology and Engineering Center, HUST's repositories
transition-aware
Transition-aware is an CRNN-based piano transcription model, inferring onset, pitch of 88 notes from raw audios. Compared with the existing models, like Onsets and Frames and High-resolution, Transition-aware is a more stable and robust approach as mentioned in our paper.
HUST_Solfege
A dataset for solfege assessment, onset detection and note transcription
AMT_project
A Two-Stage Audio-Visual Fusion Piano Transcription Model based on Attention Mechanism
Language:Python000