Rare event detection (RED) involves the identification and detection of events characterized by low frequency of occurrences, but of high importance or impact. This paper presents a Systematic Review (SR) of rare event detection across various modalities using Machine Learning (ML) and Deep Learning (DL) techniques. This review comprehensively outlines techniques and methods best suited for rare event detection across various modalities, while also highlighting future research prospects. To the extent of our knowledge, this paper is a pioneering SR dedicated to exploring this specific research domain. This SR identifies the employed methods and techniques, the datasets utilized, and the effectiveness of these methods in detecting rare events. Four modalities concerning RED are reviewed in this SR: video, sound, image, and time series. The corresponding performances for the different ML and DL techniques for RED are discussed comprehensively, together with the associated RED challenges and limitations as well as the directions for future research are highlighted. This SR aims to offer a comprehensive overview of the existing methods in RED, serving as a valuable resource for researchers and practitioners working in the respective field.

According to the World Health Organization, approximately 6 in 100 people suffer from Post-traumatic stress disorder (PTSD) at some point in their lives. PTSD is a mental disorder with a set of symptoms observed in a person following exposure to a life-threatening event or witness a death. The primary characteristic of PTSD is the persistence over time of certain symptoms: flashbacks, traumatic nightmares, acute stress and symptoms of depression. In a stressful situation, the whole body goes into tension. This has an energy cost and affects the voice, breathing, daily gestures and the face and which are referred to in this survey paper as external symptoms. The diagnosis of PTSD is based on the Diagnostic and Statistical Manual of Mental Disorders (also known by the acronym DSM) classification using standard questionnaires where the patient self-reports their condition based on their symptoms. Indeed, these questionnaires have several limitations and give an imprecise diagnosis. Sensors and wearable based technologies can play a key role in improving the diagnosis, the prognosis and the assistance of PTSD. Recently, Computer-Aided Diagnosis (CAD) systems have been proposed for PTSD detection based on external symptoms and brain activity using video and EEG sensors. To the best of our knowledge, this is the first survey paper that gives a literature overview of machine learning based approaches for PTSD diagnosis using video and EEG sensors. In addition, a comparison between existing approaches and a discussion about the potential avenues for future PTSD research is provided.