Interesting links, roughly 15/9/2021

H5P – used for learning materials on TG4 and Tuairisc

Commits on transformers: Add SpeechEncoderDecoder & Speech2Text2, Add the AudioClassificationPipeline, Add Wav2Vec2 & Hubert ForSequenceClassification (based on converting s3rpl checkpoints)

monologg/JointBERT – (Unofficial) Pytorch implementation of JointBERT: BERT for Joint Intent Classification and Slot Filling

deezer/spleeter – Deezer source separation library including pretrained models.

VoxLingua107: a Dataset for Spoken Language Recognition – no Irish

Appen/UHV-OTS-Speech – A data annotation pipeline to generate high-quality, large-scale speech datasets with machine pre-labeling and fully manual auditing. Paper

The Effects of Automatic Speech Recognition Quality on Human Transcription Latency

@inproceedings{gaur16latency,
  author    = {Yashesh Gaur and
               Walter S. Lasecki and
               Florian Metze and
               Jeffrey P. Bigham},
  editor    = {Gregory R. Gay and
               Tiago Jo{\~{a}}o Guerreiro},
  title     = {The {E}ffects of {A}utomatic {S}peech {R}ecognition {Q}uality on {H}uman {T}ranscription {L}atency},
  booktitle = {Proceedings of the 13th Web for All Conference, {W4A} '16, Montreal,
               Canada, April 11-13, 2016},
  pages     = {23&#58;1--23&#58;8},
  publisher = ,
  year      = {2016},
  doi       = {10.1145/2899475.2899478},
}