Fisher
Benchmarks
TaskDatasetSOTA resultTrendResults
Fisher (test)
7.3GPT-4o Score
64
Fisher Spanish-English (test)
90.5BLEU (Speech Input)
55
Fisher Spanish-English (dev)
88.5BLEU (Speech)
48
Fisher Spanish-English (dev2)
89.4ASR BLEU
36
Fisher (test)
9.33GPT-4o Score
32
Fisher Monolingual (test)
35.87BLEU
11
Fisher Code-Switching (test)
37.51BLEU
11
Fisher (test)
0.9WDER
11
Fisher Es→En (test)
70.2ASR chrF
10
Fisher Es→En (dev)
69.5ASR chrF
10
Fisher (test)
58Occurrence Proportion
7
Fisher (test)
60Occurrence Rate
7
Fisher
97.8Init Rate
5
Fisher
0.69Onset MAE
5
Fisher
4.25M-MOS
4
Fisher (test)
64.7BLEU
4
Fisher Global Meeting-level
15.21DER
4
Fisher (local setting)
8.18DER
4
Fisher
66.3Proportion of Correct Selections
3
Fisher
69.7Proportion Correct Selections
3
Fisher
64.76Exact Accuracy
1
Fisher (Human-Human) OOD (test)
98.44Accuracy
1