README.md

metadata

license: cc-by-4.0
task_categories:
  - text-to-speech
  - automatic-speech-recognition
language:
  - la
pretty_name: Vox Classica
tags:
  - audio
  - speech-processing
  - text-to-speech
  - tts
  - automatic-speech-recognition
  - asr
  - latin
  - classical-latin
  - low-resource

Dataset Summary

Vox Classica is a Latin speech corpus of ~67 hours of audio, segmented into samples each under 30 seconds in length. Vox Classica is a large-scale, ML-ready dataset of human-read Classical Latin. It was designed to address the absence of a publicly available human-read Latin corpus large enough for model training.

• Curated by: Kaiyuan (Ken) Zhao
• Language: Latin (Classical)
• Sources: Audio primarily derived from the excellent audio collection provided by Dickinson College Commentaries. Text primarily sourced from the public domain texts available via The Perseus Project.

Uses

This dataset is built for training and evaluating speech processing models for Classical Latin. Its primary intended use is for training and fine-tuning Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) models.

Dataset Structure

Each example in the dataset represents roughly 1-2 sentences and contains the following fields:

• transcription: A string containing the gold-standard Latin sentence. The text has been automatically macronized using the CLTK macronizer to provide correct vowel length information.
• audio: An mp3 feature containing the spoken version of the text.

Dataset Creation

To split the long-form audio recordings and the corresponding "gold standard" texts into short audio clips, the gold text was segmented into individual sentences. To approximate the location of each sentence within the long audio file, a pre-trained whisper-large-v3 model was used to generate a rough transcript of the entire audio. A fuzzy string matching algorithm was used to find the location of each gold-standard sentence within the noisy Whisper transcript, providing estimated start and end timestamps. The curator personally verified and adjusted the segment endpoints to ensure they precisely aligned with the sentence text.