cifkao commited on
Commit
2a620a4
1 Parent(s): 804df9b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -0
README.md CHANGED
@@ -80,4 +80,30 @@ compute_metrics(dataset["text"], transcriptions, languages=dataset["language"])
80
  Alternatively, if you already have transcriptions, you might prefer to skip loading the audio:
81
  ```python
82
  dataset = load_dataset("audioshake/jam-alt", revision="v1.0.0", with_audio=False)["test"]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
83
  ```
 
80
  Alternatively, if you already have transcriptions, you might prefer to skip loading the audio:
81
  ```python
82
  dataset = load_dataset("audioshake/jam-alt", revision="v1.0.0", with_audio=False)["test"]
83
+ ```
84
+
85
+ ## Citation
86
+
87
+ When using the benchmark, please cite our paper as well as the original Jamendo Lyrics paper:
88
+ ```bibtex
89
+ @misc{cifka-2023-jam-alt,
90
+ author = {Ond\v{r}ej C\'ifka and
91
+ Constantinos Dimitriou and
92
+ {Cheng-i} Wang and
93
+ Hendrik Schreiber and
94
+ Luke Miner and
95
+ Fabian-Robert St\"oter},
96
+ title = {{Jam-ALT}: A Formatting-Aware Lyrics Transcription Benchmark},
97
+ eprint = {arXiv:2311.13987},
98
+ year = 2023
99
+ }
100
+ @inproceedings{durand-2023-contrastive,
101
+ author={Durand, Simon and Stoller, Daniel and Ewert, Sebastian},
102
+ booktitle={2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
103
+ title={Contrastive Learning-Based Audio to Lyrics Alignment for Multiple Languages},
104
+ year={2023},
105
+ pages={1-5},
106
+ address={Rhodes Island, Greece},
107
+ doi={10.1109/ICASSP49357.2023.10096725}
108
+ }
109
  ```