Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

DAIC-WOZ

Benchmarks

Task NameDataset NameSOTA ResultTrend
Depression DetectionDAIC-WOZ (dev)
F1 (avg)90
43
Depression AssessmentDAIC-WOZ
Mean Absolute Error2.4434
22
Depression DetectionDAIC-WOZ
F1 Score79.53
17
Depression DetectionDAIC-WOZ (val)
Accuracy84.75
10
Depression DetectionDAIC-WOZ
Weighted F1-score68.5
8
Depression DetectionDAIC-WOZ (held-out)
F1 Score0.749
8
PHQ-8 score estimationDAIC-WOZ (test)
MAE (Stage 1)2.89
7
Depression DetectionDAIC-WOZ (cross-validation)
AUC0.689
7
Depression DetectionDAIC-WOZ
Macro F1 Score89.99
7
Sentiment Analysis (SEN)DAIC–WOZ
Binary Weighted F173.8
7
PII ExtractionDAIC-WOZ (test)
Accuracy (ACC)100
6
Audio Reconstruction FidelityDAIC-WOZ
MSE0
6
Depression screeningDAIC-WOZ (In-Sample (IS))
F1 Score68.7
5
Depression classificationDAIC-WoZ AVEC 2016 benchmark
Accuracy0.914
5
Depression DetectionDAIC-WOZ (train)
Accuracy79
4
Causal Utterance RankingDAIC-WOZ (test)
Hit@146.1
4
Depression classificationDAIC-WoZ (K-fold LOSO)
Accuracy0.957
4
Depression DetectionDAIC-WOZ (test)
Weighted Accuracy71.4
2
Showing 18 of 18 rows