Lahjoita puhetta semisupervised baseline speech recognition model, built with the Kaldi toolkit. Trained on 100 hours of supervised and approx. 1600 hours of untranscribed Finnish speech. Described in more detail in the paper https://arxiv.org/abs/2203.12906 "Lahjoita puhetta – a large-scale corpus of spoken Finnish with some benchmarks". For details on the training method, see https://github.com/aalto-speech/lahjoita-puhetta-baseline-kaldi. The title and description of this software/code correspond with the situation when the software metadata was imported to ACRIS. The most recent version of metadata is available in the original repository.

2022