Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages · HackerPapers discussions

Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages — mentioned by lysozyme

Thu, 30 Mar 2023 07:41:00 GMT

Interesting that they enriched the training data by asking people to point out YouTube videos in specific languages for which they needed data > YT-513-U: We create an additional dataset called YT-513-U to ensure coverage of lower resource languages in our pre-training dataset. We reached out to vendors and native speakers to identify YT videos containing speech in specific long tail languages, collecting a dataset of unlabeled speech in 513 languages. [1] 1. https://arxiv.org/abs/2303.01037

Paper and discussion history. A mention is not an endorsement.

Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages — mentioned by runnerup

Thu, 30 Mar 2023 01:08:19 GMT

In the paper itself[0], they do have a couple comparisons for English-only (Fig. 2, Table 3), and it looks like Whisper has an 11.5% error rate and USM has a 10.5% error rate. It's a truly negligible difference. There's no way I'd ever pay for an API for this if I knew I only cared about English. I know that's not the point of this model (the point is that for a lot of languages, its the only model available). But paywalling it seems greedy, you'll only extract money from those under-represented communities. On the other hand, maybe this never would have been built without the profit motive.…

Paper and discussion history. A mention is not an endorsement.

Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages — HN submission

Tue, 07 Mar 2023 13:10:05 GMT

Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.