Papers › Scribosermo: Fast Speech-to-Text models for German and other Languages

Scribosermo: Fast Speech-to-Text models for German and other Languages

15 Oct 2021arXiv:2110.07982archive 2025-07-28

Daniel Bermuth, Alexander Poeppel, Wolfgang Reif

Recent Speech-to-Text models often require a large amount of hardware resources and are mostly trained in English. This paper presents Speech-to-Text models for German, as well as for Spanish and French with special features: (a) They are small and run in real-time on microcontrollers like a RaspberryPi. (b) Using a pretrained English model, they can be trained on consumer-grade hardware with a relatively small dataset. (c) The models are competitive with other solutions and outperform them in German. In this respect, the models combine advantages of other approaches, which only include a subset of the presented features. Furthermore, the paper provides a new library for handling datasets, which is focused on easy extension with additional datasets and shows an optimized way for transfer-learning new languages using a pretrained model from another language with a similar alphabet.

PaperPDFCode

Code

gitlab.com/jaco-assistant/scribosermo officialmentioned in papermentioned on GitHubtf report
gitlab.com/jaco-assistant/corcua officialmentioned in paper report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Speech RecognitionSpeech-to-TextTransfer Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Speech Recognition Common Voice French ConformerCTC-L (5-gram) Test WER 8.13% #1 of 8 Archive leaderboard report
Speech Recognition Common Voice French ConformerCTC-L (no-LM) Test WER 10.19 % #5 of 8 Archive leaderboard report
Speech Recognition Common Voice French QuartzNet15x5FR (D7) Test WER 11.0% #6 of 8 Archive leaderboard report
Speech Recognition Common Voice French QuartzNet15x5FR (CV-only) Test WER 12.1% #7 of 8 Archive leaderboard report
Speech Recognition Common Voice German ConformerCTC-L (5-gram) Test CER 1.37% #3 of 14 Archive leaderboard report
Speech Recognition Common Voice German ConformerCTC-L (5-gram) Test WER 4.05% #3 of 14 Archive leaderboard report
Speech Recognition Common Voice German QuartzNet15x5DE (D37, 5-gram) Test CER 2.7% #8 of 14 Archive leaderboard report
Speech Recognition Common Voice German QuartzNet15x5DE (D37, 5-gram) Test WER 6.6% #8 of 14 Archive leaderboard report
Speech Recognition Common Voice German ConformerCTC-L (no LM) Test CER 2.05% #10 of 14 Archive leaderboard report
Speech Recognition Common Voice German ConformerCTC-L (no LM) Test WER 7.33% #10 of 14 Archive leaderboard report
Speech Recognition Common Voice German QuartzNet15x5DE (CV-only, 5-gram) Test CER 3.2% #11 of 14 Archive leaderboard report
Speech Recognition Common Voice German QuartzNet15x5DE (CV-only, 5-gram) Test WER 7.7% #11 of 14 Archive leaderboard report
Speech Recognition Common Voice Italian QuartzNet15x5IT (D5) Test WER 11.5% #2 of 2 Archive leaderboard report
Speech Recognition Common Voice Spanish ConformerCTC-L (5-gram) Test WER 5.68% #3 of 8 Archive leaderboard report
Speech Recognition Common Voice Spanish ConformerCTC-L (no-LM) Test WER 7.46 % #5 of 8 Archive leaderboard report
Speech Recognition Common Voice Spanish QuartzNet15x5ES (D8) Test WER 10.0% #6 of 8 Archive leaderboard report
Speech Recognition Common Voice Spanish QuartzNet15x5ES (CV-only) Test WER 10.5% #8 of 8 Archive leaderboard report
Speech Recognition TUDA QuartzNet15x5DE (D37) Test WER 10.2% #3 of 9 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections