Abstract: A speech recognition module receives training data of speech and creates a representation for individual words, non-words, phonemes, and any combination. A set of speech processing detectors analyze the training data of speech from humans communicating. The set of speech processing detectors detect speech parameters that are indicative of paralinguistic effects on top of enunciated words, phonemes, and non-words in the audio stream. One or more machine learning models undergo supervised machine learning on their neural network to train on how to associate one or more mark-up markers with a textual representation, for each individual word, individual non-word, individual phoneme, and any combinations of these, that was enunciated with a particular paralinguistic effect. Each mark-up marker can correspond to its own paralinguistic effect.

Abstract: A computer-implemented method can include a speech collection module collecting a speech pattern from a patient, a speech feature computation module computing at least one speech feature from the collected speech pattern, a mental health determination module determining a state-of-mind of the patient based at least in part on the at least one computed speech feature, and an output module providing an indication of a diagnosis with regard to a possibility that the patient is suffering from a certain condition such as depression or Post-Traumatic Stress Disorder (PTSD).

Abstract: A computer-implemented method can include a speech collection module collecting a speech pattern from a patient, a speech feature computation module computing at least one speech feature from the collected speech pattern, a mental health determination module determining a state-of-mind of the patient based at least in part on the at least one computed speech feature, and an output module providing an indication of a diagnosis with regard to a possibility that the patient is suffering from a certain condition such as depression or Post-Traumatic Stress Disorder (PTSD).

Abstract: A method for classifying lexical stress in an utterance includes generating a feature vector representing stress characteristics of a syllable occurring in the utterance, wherein the feature vector includes a plurality of features based on prosodic information and spectral information, computing a plurality of scores, wherein each of the plurality of scores is related to a probability of a given class of lexical stress, and classifying the lexical stress of the syllable based on the plurality of scores.

Abstract: Technologies for analyzing temporal components of multimodal data to detect short-term multimodal events, determine relationships between short-term multimodal events, and recognize long-term multimodal events, using a deep learning architecture, are disclosed.

Abstract: Technologies for analyzing temporal components of multimodal data to detect short-term multimodal events, determine relationships between short-term multimodal events, and recognize long-term multimodal events, using a deep learning architecture, are disclosed.

Abstract: A method for classifying lexical stress in an utterance includes generating a feature vector representing stress characteristics of a syllable occurring in the utterance, wherein the feature vector includes a plurality of features based on prosodic information and spectral information, computing a plurality of scores, wherein each of the plurality of scores is related to a probability of a given class of lexical stress, and classifying the lexical stress of the syllable based on the plurality of scores.