Comparing Emotion Recognition and Word Recognition in Background Noise

Abstract

Purpose Word recognition in quiet and in background noise has been thoroughly investigated in previous research to establish segmental speech recognition performance as a function of stimulus characteristics (e.g., audibility). Similar methods to investigate recognition performance for suprasegmental information (e.g., acoustic cues used to make judgments of talker age, sex, or emotional state) have not been performed. In this work, we directly compared emotion and word recognition performance in different levels of background noise to identify psychoacoustic properties of emotion recognition (globally and for specific emotion categories) relative to word recognition. Method Twenty young adult listeners with normal hearing listened to sentences and either reported a target word in each sentence or selected the emotion of the talker from a list of options (angry, calm, happy, and sad) at four signal-to-noise ratios in a background of white noise. Psychometric functions were fit to the recognition data and used to estimate thresholds (midway points on the function) and slopes for word and emotion recognition. Results Thresholds for emotion recognition were approximately 10 dB better than word recognition thresholds, and slopes for emotion recognition were half of those measured for word recognition. Low-arousal emotions had poorer thresholds and shallower slopes than high-arousal emotions, suggesting greater confusion when distinguishing low-arousal emotional speech content. Conclusions Communication of a talker's emotional state continues to be perceptible to listeners in competitive listening environments, even after words are rendered inaudible. The arousal of emotional speech affects listeners' ability to discriminate between emotion categories.

Age-Related Changes in the Perception of Emotions in Speech: Assessing Thresholds of Prosody and Semantics Recognition in Noise for Young and Older Adults

“…This suggests that to reach recognition threshold in noise, emotional semantics call for an addition of ∼3.3 dB SNR as compared to emotional prosody. This prosodic advantage across age groups expands previous evidence that focused mainly on an accuracy advantage for prosodic recognition over spoken word recognition ( Dupuis and Pichora-Fuller, 2014 ; Morgan, 2021 ; Morgan et al, 2022 ). A noteworthy study by Morgan (2021) reported a 10 dB SNR advantage between emotional prosodic thresholds and spoken word identification thresholds in young adults.…”

“…This prosodic advantage across age groups expands previous evidence that focused mainly on an accuracy advantage for prosodic recognition over spoken word recognition ( Dupuis and Pichora-Fuller, 2014 ; Morgan, 2021 ; Morgan et al, 2022 ). A noteworthy study by Morgan (2021) reported a 10 dB SNR advantage between emotional prosodic thresholds and spoken word identification thresholds in young adults. This marks a much larger advantage than the 3.3 dB SNR difference we report.…”

“…To date, only a few studies have tried to directly compare the recognition of the semantic and prosodic channels, mostly in young adults (but see, Dupuis and Pichora-Fuller, 2014 ). For example, Morgan (2021) showed that sensory thresholds were better for prosodic-emotion recognition than for word recognition in noise for young adults (see also van Zyl and Hanekom, 2011 ; Ritter and Vongpaisal, 2018 ; Morgan et al, 2022 ). However, these studies did not directly measure semantic-emotion recognition, but rather used word/sentence recognition as a placeholder.…”

“…The crossover point is taken to represent the point at which the rate of increase in recognition as a function of SNR begins to decrease. As such, the value of x 0 can serve as an index for individual recognition statistical threshold ( Ben-David et al, 2012 ; Morgan, 2021 ). Finally, k represents the function’s slope at x 0 .…”

“…(2) We added an estimation level of 0.2 recognition rates (chance level) for an SNR of −20 dB, to correspond to the function’s predicted lower bound. However, we chose not to pre-define the upper bound of the function (i.e., maximum recognition rates, see Morgan, 2021 ), as even without any background noise emotional recognition rates are not expected to reach 100%, especially for older adults (see Ruffman et al, 2008 ; Ben-David et al, 2019 ). Hence, the three other parameters ( x 0 , L , and k ) were estimated based on our data.…”