speakerbox 0.0.2

Speaker Annotation for Transcripts using Audio Classification

speakerbox

Speaker Annotation for Transcripts using Audio Classification

---

Installation

Stable Release: pip install speakerbox
Development Head: pip install git+https://github.com/CouncilDataProject/speakerbox.git

Documentation

For full package documentation please visit councildataproject.github.io/speakerbox.

Quickstart

Load the 2021 Seattle Prototype Dataset, get summary statistics about speaker time, finally pull the matching audio file for each annotation file and store annotation file matched to audio as a pandas.DataFrame.

from speakerbox import datasets

seattle_2021_ds_dir = datasets.unpack_seattle_2021_proto(clean=True)
seattle_2021_ds_summary_stats = datasets.summarize_annotation_statistics(
    seattle_2021_ds_dir / "annotations"
)
seattle_2021_ds = datasets.pull_seattle_2021_proto_audio()

Development

See CONTRIBUTING.md for information related to developing the code.

MIT license