DAIC-WOZ
Evaluation was internal to DAIC-WOZ, so external validation is needed before generalization. The dataset was modest in size and class-imbalanced. The analyzed labeled subset included 142 participants with binary PHQ-8 depression labels. Th…
1 sources - 4 claims
Evaluation was internal to DAIC-WOZ, so external validation is needed before generalization. The dataset was modest in size and class-imbalanced. The analyzed labeled subset included 142 participants with binary PHQ-8 depression labels. The study used the depression subset of DAIC-WOZ, a clinical interview corpus based on human-computer interactions and virtual interviewer technology.