Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Device-directed speech classification on Internal (held-out set)
Loading...
95
F1 Score
SAS
85.64
88.07
90.5
92.93
Apr 9, 2026
F1 Score
Precision
Recall
Worst-case Session F1 Score
Average Precision
False Trigger Rate
Decision Latency (ms)
Runtime Footprint (MB)
Updated 9d ago
Evaluation Results
Method
Method
Links
F1 Score
Precision
Recall
Worst-case Session F1 Score
Average Precision
False Trigger Rate
Decision Latency (ms)
Runtime Footprint (MB)
SAS
Input modality=audio+v...
2026.04
95
97
93
88
87
-
150
20
SAS
Input modality=audio-o...
2026.04
86
89
83
88
87
2.1
150
20
SAS
Input modality=audio-o...
2026.04
-
-
-
-
-
7.8
-
-
Feedback
Search any
task
Search any
task
