A multimodal model that uses DASS-21's item structure and LLM summaries improves depression/anxiety/stress risk and item prediction on the AdoDAS validation split.
The Verbal and Non Verbal Signals of Depression -- Combining Acoustics, Text and Visuals for Estimating Depression Level
1 Pith paper cite this work, alongside 16 external citations. Polarity classification is still indexing.
1
Pith paper citing it
16
external citations · Pith
abstract
Depression is a serious medical condition that is suffered by a large number of people around the world. It significantly affects the way one feels, causing a persistent lowering of mood. In this paper, we propose a novel attention-based deep neural network which facilitates the fusion of various modalities. We use this network to regress the depression level. Acoustic, text and visual modalities have been used to train our proposed network. Various experiments have been carried out on the benchmark dataset, namely, Distress Analysis Interview Corpus - a Wizard of Oz (DAIC-WOZ). From the results, we empirically justify that the fusion of all three modalities helps in giving the most accurate estimation of depression level. Our proposed approach outperforms the state-of-the-art by 7.17% on root mean squared error (RMSE) and 8.08% on mean absolute error (MAE).
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DynaBridge: Dynamic Summary-Guided Cross-Task Multimodal Fusion for DASS-Structured Mental Health Assessment
A multimodal model that uses DASS-21's item structure and LLM summaries improves depression/anxiety/stress risk and item prediction on the AdoDAS validation split.