In line with the theme for INTERSPEECH 2018 (Speech Research for Emerging Markets in Multilingual Societies),
the emphasis is on (but not limited to) the below topics, arranged according to the workshop program.

Detecting stress, emotion or mental states of people from speech

  • Non-acted speech databases
  • Multiple languages (e.g. Indian languages)
  • Realistic environments (e.g. call centers, workplaces)
  • Use of other modes such as video and sensor data, preferably from wearable sensors, in addition to speech

Effects of Audio on stress, emotion and mental states of people

  • Diverse music such as folk, classical and film genres
  • The dependency on cultural backgrounds is of special interest
  • Use of audio cues to influence mental states / human behavior

Other topics that are of interest in the context of stress, emotion and mental states

  • Novel signal processing or machine learning techniques
  • Sounds at inaudible frequencies
  • Novel protocols for assessing mental states, inducing stress or emotion
  • Applications related to the above topics

Submission Process

Submitted papers were reviewed by the Scientific Committee. Each paper received at least two reviews.
The list of accepted papers is below. Authors have been requested to register so that they can be presented orally at the workshop.

Accepted Papers

  • Time-frequency spectral error for analysis of high arousal speech
  • Task-Independent EEG based Subject Identification using Auditory Stimulus
  • Speech Emotion Recognition with Data Augmentation and Layer-wise Optimizer Adjustment
  • Emotional Speech Classifier Systems: For Sensitive Assistance to support Disabled Individuals
  • Discriminating between High-Arousal and Low-Arousal Emotional States of Mind using Acoustic Analysis
  • Analysis of Speech Emotions in Realistic Environments
  • A component-based approach to study the impact of Indian music on emotions