[ASR] Online Speaker Diarization with ASR: Part-1

[tango4j]

What does this PR do ?

This PR addresses utility functions (not core functions) for online diarization with ASR (multi-speaker ASR).
Streaming ASR and diarization utility functions and classes are added with unit tests.

This is Part1 of the clustering online diarization integration.
In Part2, example file, yaml file and tutorial notebook will be uploaded.

Collection: ASR

Changelog

nemo/collections/asr/models/__init__.py
Adding OnlineClusteringDiarizer class

nemo/collections/asr/parts/utils/decoder_timestamps_utils.py
adding FrameBatchASRLogitsSample and FrameBatchASRLogits
for ASR with timestamp generation

nemo/collections/asr/parts/utils/diarization_utils.py
OnlineDiarWithASR and a few utility functions

nemo/collections/asr/parts/utils/streaming_utils.py
Adding FrameBatchVAD class which handles VAD streaming.

Usage

Usage will be included in part2 with example file and yaml file.

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation?
• [] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
• [] Reviewer: Does the PR have correct import guards for all optional libraries?

PR Type:

• New Feature
• Bugfix
• Documentation

If you haven't finished some of the above items you can still open "Draft" PR.

Who can review?

Anyone in NeMo ASR.

Additional Information

N.A.

[github-advanced-security]

CodeQL found more than 10 potential problems in the proposed changes. Check the Files changed tab for more details.

[github-actions]

This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.