[REQ] Audio-to-MIDI transcription models

Hi there,
it would be great to exploit these models too:
- Spotify's (@psobot) [Basic Pitch](https://github.com/spotify/basic-pitch-ts#readme) ([bin](https://github.com/spotify/basic-pitch-ts/tree/main/model));
- Google's (@iansimon) [MT3: Multi-Task Multitrack Music Transcription](https://github.com/magenta/mt3#readme) ([gins](https://github.com/magenta/mt3/tree/main/mt3/gin))
- MCTLab's (@BreezeWhite) [OMNIZART](https://github.com/Music-and-Culture-Technology-Lab/omnizart#readme) ([checkpoints](https://github.com/Music-and-Culture-Technology-Lab/omnizart/tree/master/omnizart/checkpoints))
- @DamRsn's [NeuralNote](https://github.com/DamRsn/NeuralNote#readme) ([ONNX](https://github.com/DamRsn/NeuralNote/tree/master/Lib/ModelData));
- @xavriley's [MIDI Transcription Model](https://github.com/xavriley/hf_midi_transcription#readme) ([pths](https://huggingface.co/xavriley/midi-transcription-models/tree/main));
- @Nkcemeka's [Snap2MIDI](https://github.com/Nkcemeka/Snap2MIDI#readme) ([pts](https://huggingface.co/nkcemeka/Snap2MIDI/tree/main)).

Hope that inspires.

Review the linked Basic Pitch, MT3, OMNIZART, NeuralNote, MIDI Transcription Model, and Snap2MIDI repositories and their referenced model files. Determine which models and integration scope this project should support before estimating the work. Done means the selected audio-to-MIDI models are supported, with behavior verified in the project.

Written by the indexing model from the issue text.