Transferring speech-generic and depression-specific knowledge for Alzheimer's disease detection

TL;DR

This paper introduces a novel knowledge transfer framework leveraging speech-generic and depression-specific models to improve Alzheimer's disease detection from speech, achieving state-of-the-art results on the ADReSSo dataset.

Contribution

It proposes a joint knowledge transfer approach from foundation models and depression detection to enhance AD diagnosis from speech data.

Findings

Improved AD detection accuracy with a state-of-the-art F1 score of 0.928.

Demonstrated effectiveness of combining speech-generic and depression-specific knowledge.

Validated the approach on the ADReSSo dataset with significant performance gains.

Abstract

The detection of Alzheimer's disease (AD) from spontaneous speech has attracted increasing attention while the sparsity of training data remains an important issue. This paper handles the issue by knowledge transfer, specifically from both speech-generic and depression-specific knowledge. The paper first studies sequential knowledge transfer from generic foundation models pretrained on large amounts of speech and text data. A block-wise analysis is performed for AD diagnosis based on the representations extracted from different intermediate blocks of different foundation models. Apart from the knowledge from speech-generic representations, this paper also proposes to simultaneously transfer the knowledge from a speech depression detection task based on the high comorbidity rates of depression and AD. A parallel knowledge transfer framework is studied that jointly learns the information shared between these two tasks. Experimental results show that the proposed method improves AD and depression detection, and produces a state-of-the-art F1 score of 0.928 for AD diagnosis on the commonly used ADReSSo dataset.