diart

diart copied to clipboard

## Problem There's been some interest in adding compatibility with networking protocols (see #9 #27 #67 #68). ## Idea Add a `TCPAudioSource` and `UDPAudioSource` following a similar implementation to `MicrophoneAudioSource`.

juanmc2005

## Problem There's been some interest in adding compatibility with networking protocols (see #9 #27 #67 #68). ## Idea Add a `WebsocketAudioSource` following a similar implementation to `MicrophoneAudioSource`.

juanmc2005

Colab example run

1

Basic Colab example notebook, can be updated and modified by you as you like. https://colab.research.google.com/drive/1Bp4_v5SNU1anDDEDkvA1FLnqGIfxM7ov?usp=sharing

AMITKESARI2000

## Problem It's difficult to understand the shape and format expected from tensors. ## Idea Add [torchtyping](https://github.com/patrick-kidger/torchtyping). There's also the possibility to check that types are respected at runtime (instead...

juanmc2005

Optionally relax incremental clustering constraints

2

## Problem Cannot-link constraints are currently hard-coded in `OnlineSpeakerClustering`. If a segmentation model over-segments speakers, it may be better to rely on speaker embeddings instead to determine the identity of...

juanmc2005

There is some dead code inside `OnlineSpeakerClustering`. It would be better to remove it to improve the quality and readability of the code base.

juanmc2005

How to add ASR

1

I've been trying a while to get streaming diarization + recognition working together. I've failed trying many different strategies for streaming wav to pyannote. I started with an example that...

bitnom

It would be nice to make diart compatible with the `torchaudio` streaming API. This would also make it possible to easily add other audio sources like the screen, videos, and...

juanmc2005

Gradio/Streamlit demo on Huggingface Space

2

_just putting it here as a note to myself :)_ * Looks like [it is now possible](https://gradio.app/real_time_speech_recognition/) to do live speech processing in [Gradio](https://gradio.app/). * There has also been [some](https://github.com/dataprofessor/realtime-transcription)...

hbredin

Paper at JOSS

11

Hi @zaouk, @Yagna24, @ckliao-nccu, @igordertigor, @AMITKESARI2000! I'm planning to submit a paper for diart at [The Journal of Open Source Software (JOSS)](https://joss.theoj.org/) and it seems natural that all of you,...

juanmc2005