Singing Voice Synthesis Task for ESPnet

[ftshijt]

This is the singing voice synthesis task for ESPnet, as noted in #4437

For now, we have supported a sample recipe and the corresponding setups (including RNN and RNN-DP models).

Some major updates from original Muskits:

• We replace the raw music score from MIDI to MusicXML for easier usage in various cases
• We fixed some legacy issues with Muskits

It would be great if you could have some suggestions and comments.

We intend to distribute the work after this initial one. Therefore, we would probably add test functions in the following PR.

[codecov]

Codecov Report

Merging #4670 (249c8e8) into master (6f61bc7) will decrease coverage by 2.64%.
The diff coverage is 2.82%.

@@            Coverage Diff             @@
##           master    #4670      +/-   ##
==========================================
- Coverage   83.09%   80.45%   -2.65%     
==========================================
  Files         518      527       +9     
  Lines       44700    46218    +1518     
==========================================
+ Hits        37143    37184      +41     
- Misses       7557     9034    +1477     

Flag	Coverage Δ
test_integration_espnet1	66.37% <ø> (ø)
test_integration_espnet2	49.05% <16.09%> (-0.31%)	⬇️
test_python	68.67% <2.82%> (-2.25%)	⬇️
test_utils	23.30% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
espnet2/bin/svs_inference.py	0.00% <0.00%> (ø)
espnet2/bin/svs_train.py	0.00% <0.00%> (ø)
espnet2/svs/abs_svs.py	0.00% <0.00%> (ø)
espnet2/svs/espnet_model.py	0.00% <0.00%> (ø)
espnet2/svs/feats_extract/score_feats_extract.py	0.00% <0.00%> (ø)
espnet2/svs/naive_rnn/naive_rnn.py	0.00% <0.00%> (ø)
espnet2/svs/naive_rnn/naive_rnn_dp.py	0.00% <0.00%> (ø)
espnet2/tasks/svs.py	0.00% <0.00%> (ø)
espnet2/train/preprocessor.py	28.69% <2.30%> (-7.97%)	⬇️
espnet2/fileio/read_text.py	72.22% <7.69%> (-20.47%)	⬇️
... and 4 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[ftshijt]

The CI seems to be failed due to some flack8 issues

[ftshijt]

Please let us know if the PR looks good~ It would be great to merge for further distribution of works.

[ftshijt]

As this is quite a dependency for following works, I would want to merge it for now~ There are plenty of TODOs to make it better. Please feel free to provide any comments if you have any~