jfreiwa commited on
Commit
f778c09
1 Parent(s): 2f34cd0

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +87 -0
README.md CHANGED
@@ -1,3 +1,90 @@
1
  ---
2
  license: cc-by-sa-4.0
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: cc-by-sa-4.0
3
+ language: "de"
4
+ thumbnail:
5
+ tags:
6
+ - automatic-speech-recognition
7
+ - CTC
8
+ - Attention
9
+ - pytorch
10
+ - speechbrain
11
+ metrics:
12
+ - wer
13
  ---
14
+
15
+ # German ASR
16
+
17
+ This model is trained on the Mozilla Common Voice 8.0, the Spoken Wikipedia Corpus and the m-ailabs corpus.
18
+
19
+ - https://nats.gitlab.io/swc/
20
+ - https://commonvoice.mozilla.org/de/datasets
21
+ - https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset/
22
+
23
+ We do not provide a language model.
24
+
25
+ # Performance
26
+
27
+ This model has a WER of 7.24%.
28
+
29
+ # Model application
30
+
31
+ ## Install SpeechBrain
32
+ First of all, please install SpeechBrain with the following command:
33
+ ```
34
+ pip install speechbrain
35
+ ```
36
+ Please notice that we encourage you to read the tutorials and learn more about
37
+ [SpeechBrain](https://speechbrain.github.io).
38
+
39
+ ## Using the model
40
+ ```
41
+ from speechbrain.pretrained import EncoderDecoderASR
42
+
43
+ asr_model = EncoderDecoderASR.from_hparams(source="jfreiwa/asr-crdnn-german", savedir="pretrained_models/asr-crdnn-german")
44
+ asr_model.transcribe_file("jfreiwa/asr-crdnn-german/example-de.wav")
45
+
46
+ ```
47
+
48
+ ## Inference on GPU
49
+ To perform inference on the GPU, add `run_opts={"device":"cuda"}` when calling the `from_hparams` method.
50
+
51
+
52
+ # Limitations
53
+ We do not provide any warranty on the performance achieved by this model when used on other datasets.
54
+
55
+ # **About SpeechBrain**
56
+ - Website: https://speechbrain.github.io/
57
+ - Code: https://github.com/speechbrain/speechbrain/
58
+ - HuggingFace: https://huggingface.co/speechbrain/
59
+
60
+ # **Citing SpeechBrain**
61
+ Please, cite SpeechBrain if you use it for your research or business.
62
+ ```bibtex
63
+ @misc{speechbrain,
64
+ title={{SpeechBrain}: A General-Purpose Speech Toolkit},
65
+ author={Mirco Ravanelli and Titouan Parcollet and Peter Plantinga and Aku Rouhe and Samuele Cornell and Loren Lugosch and Cem Subakan and Nauman Dawalatabad and Abdelwahab Heba and Jianyuan Zhong and Ju-Chieh Chou and Sung-Lin Yeh and Szu-Wei Fu and Chien-Feng Liao and Elena Rastorgueva and François Grondin and William Aris and Hwidong Na and Yan Gao and Renato De Mori and Yoshua Bengio},
66
+ year={2021},
67
+ eprint={2106.04624},
68
+ archivePrefix={arXiv},
69
+ primaryClass={eess.AS},
70
+ note={arXiv:2106.04624}
71
+ }
72
+ ```
73
+
74
+ # **Citing our paper**
75
+ Please, cite our paper, when you use this model in your research.
76
+ ```bibtex
77
+ @inproceedings{freiwald2022,
78
+ author={J. Freiwald and P. Pracht and S. Gergen and D. Kolossa},
79
+ title={Open-Source End-To-End Learning for Privacy-Preserving German {ASR}},
80
+ year=2022,
81
+ booktitle={DAGA 2022}
82
+ }
83
+ ```
84
+
85
+ # Acknowledgements
86
+ This work was funded by the German Federal Ministry of Education and Research (BMBF)
87
+ within the “Innovations for Tomorrow’s Production, Services, and
88
+ Work” Program (02L19C200), a project that is implemented by
89
+ the Project Management Agency Karlsruhe (PTKA). The authors
90
+ are responsible for the content of this publication.