Depression is a psychological disorder that affects both the mental and physical health of the human body. Detecting the depression state of an individual provides the severity of depression symptoms and safeguard the sufferer from committing suicide. Early detection is important for depression prevention which potentially reduces the intensification of the disorder. Early depression detection aims to identify sufferers showing signs of depression that suggest appropriate diagnosis and treatment at the early stages.
Deep learning technology affords a feasible tool to identify the symptoms of depression that enables quick preventive involvement. However, Deep learning models suffer from multi-modal data sufficiency problems and lack of intelligible reason for depressed state. Deep learning with attention mechanism possesses the ability to tackle such problems by multi-modal knowledge-attention representation to classify and predict depression information. Early depression detection with a deep attention network produces a robust and effective depression detection model with superior performance.