Automatic depression diagnosis is a challenging problem, that requires integrating spatial-temporal information and extracting features from audio-visual signals. In terms of privacy protection, the development trend of recognition algorithms based on facial landmarks has created additional challenges and difficulties. In this paper, we propose an audio-visual attention network (AVA-DepressNet) for depression recognition. It is a novel multimodal framework with facial privacy protection, and uses attention-based modules to enhance audio-visual spatial and temporal features. In addition, an adversarial multistage (AMS) training strategy is developed to optimize the encoder-decoder structure. Additionally, facial structure prior knowledge is creatively used in AMS training. Our AVA-DepressNet is evaluated on popular audio-visual depression datasets: AVEC 2013, AVEC 2014, and AVEC 2017. The results show that our approach reaches the state-of-the-art performance or competitive results for depression recognition.

Recent visual-based depression recognition methods mostly use hand-crafted features with information lost in color channels, or deep network features with a limited performance from the finite data. In this paper, we propose a method called Local Quaternion and Global Deep Network (LQGDNet) which can combine advantages from hand-crafted and deep features. Specifically, the Quaternion XOR Asymmetrical Regional Local Gradient Coding (XOR-AR-LGC) is firstly designed, which encodes the facial images with local textures in the quaternion domain to keep the dependence of color channels, and integrated into the Quaternion Feature Extractor (QFE). To the best of our knowledge, it is the first attempt to use a quaternion-based method for facial depression recognition. Second, we design the Local Quaternion Representation Module (LQRM) composed of Local Deep Feature Extractor (LDFE) and QFE to output local quaternion facial features. Third, global deep facial features are encoded from the Global Deep Representation Module (GDRM) with the deep convolutional neural network. Finally, the LQGDNet integrates LQRM and GDRM with the local quaternion and global deep features and predicts the depression score. The experimental results on AVEC 2013 and AVEC 2014 show the superiority of our method compared to the state-of-the-art approaches.

Related studies have revealed that the phonological features of depressed patients are different from those of healthy individuals. With the increasing prevalence of depression, objective and convenient early screening is necessary. To this end, we propose an automatic depression detection method based on hybrid speech features extracted by deep learning, dubbed as TTFNet. Firstly, to effectively excavate the intrinsic relationship among multidimensional dynamic features in the frequency domain, the Mel spectrogram of raw speech and its related derivatives are encoded into quaternion representation. Then, the innovatively designed quaternion VisionLSTM is utilized to capture their synergistic effects. Simultaneously, we integrate sLSTM with the pre-trained wav2vec 2.0 model to fully acquire the temporal features. In addition, to further exploit the complementarity between temporal and frequency features, we design an XConformer block for cross-sequence interactions, which ingeniously combines self-attention mechanisms and convolutional modules. The designed XCFF fusion module, based on the XConformer block, enables multi-level interactions between frequency-domain and temporal-domain, thereby enhancing generalization ability of the proposed model. Extensive experiments conducted on the AVEC 2013, AVEC 2014, DAIC-WOZ and E-DAIC datasets demonstrate that our method outperforms current state-of-the-art methods in both depression recognition and severity prediction tasks.