The increasing use of computer-generated speech in various applications has raised questions about how people perceive synthetic voices. This study investigates the role of linguistic information in the perception of humanness in speech. We conducted two experiments with native German-, Spanish- and Turkish-speaking participants who rated the human-likeness of human and text-to-speech (TTS)-generated voices. By presenting German sentences as well as manipulated versions of those sentences in terms of syntax and semantics, we examined the role of three types of linguistic information, that is, phonetics, semantics, and syntax, on humanness perception. Acoustic analyses revealed differences between human and TTS-generated voices in terms of summary acoustics and dynamic contours of pitch and intensity, thus showing that TTS-generated voices are not yet fully aligned with human voices on voice quality and prosody. Importantly, behavioral results showed that these acoustic differences were more salient to native German listeners, who distinguished between human and synthetic voices more extremely. In addition to the role of phonetic or phonological familiarity, we observed a role of both syntax and semantics in humanness perception, with the manipulated sentences sounding less human regardless of the speaker (i.e., TTS-generated or human), but only for the native speakers. Lastly, humanness perception of speech appears to be relatively idiosyncratic as indicated by the individual differences observed. Altogether, this study contributes to our understanding of the interplay between linguistic and paralinguistic information in speech perception, and clarifies how listeners perceive their increasingly synthetically-generated soundscape.

The ACM Multimedia 2022 Computational Paralinguistics Challenge addresses four different problems for the first time in a research competition under well-defined conditions: In the Vocalisations and Stuttering Sub-Challenges, a classification on human non-verbal vocalisations and speech has to be made; the Activity Sub-Challenge aims at beyond-audio human activity recognition from smartwatch sensor data; and in the Mosquitoes Sub-Challenge, mosquitoes need to be detected. We describe the Sub-Challenges, baseline feature extraction, and classifiers based on the 'usual' ComParE and BoAW features, the auDeep toolkit, and deep feature extraction from pre-trained CNNs using the DeepSpectrum toolkit; in addition, we add end-to-end sequential modelling, and a log-mel-128-BNN.

Recent TTS systems are able to generate prosodically varied and realistic speech. However, it is unclear how this prosodic variation contributes to the perception of speakers' emotional states. Here we use the recent psychological paradigm 'Gibbs Sampling with People' to search the prosodic latent space in a trained GST Tacotron model to explore prototypes of emotional prosody. Participants are recruited online and collectively manipulate the latent space of the generative speech model in a sequentially adaptive way so that the stimulus presented to one group of participants is determined by the response of the previous groups. We demonstrate that (1) particular regions of the model's latent space are reliably associated with particular emotions, (2) the resulting emotional prototypes are well-recognized by a separate group of human raters, and (3) these emotional prototypes can be effectively transferred to new sentences. Collectively, these experiments demonstrate a novel approach to the understanding of emotional speech by providing a tool to explore the relation between the latent space of generative models and human semantics.

A core problem in cognitive science and machine learning is to understand how humans derive semantic representations from perceptual objects, such as color from an apple, pleasantness from a musical chord, or seriousness from a face. Markov Chain Monte Carlo with People (MCMCP) is a prominent method for studying such representations, in which participants are presented with binary choice trials constructed such that the decisions follow a Markov Chain Monte Carlo acceptance rule. However, while MCMCP has strong asymptotic properties, its binary choice paradigm generates relatively little information per trial, and its local proposal function makes it slow to explore the parameter space and find the modes of the distribution. Here we therefore generalize MCMCP to a continuous-sampling paradigm, where in each iteration the participant uses a slider to continuously manipulate a single stimulus dimension to optimize a given criterion such as ‘pleasantness’. We formulate both methods from a utility-theory perspective, and show that the new method can be interpreted as ‘Gibbs Sampling with People’ (GSP). Further, we introduce an aggregation parameter to the transition step, and show that this parameter can be manipulated to flexibly shift between Gibbs sampling and deterministic optimization. In an initial study, we show GSP clearly outperforming MCMCP; we then show that GSP provides novel and interpretable results in three other domains, namely musical chords, vocal emotions, and faces. We validate these results through large-scale perceptual rating experiments. The final experiments use GSP to navigate the latent space of a state-of-the-art image synthesis network (StyleGAN), a promising approach for applying GSP to high-dimensional perceptual spaces. We conclude by discussing future cognitive applications and ethical implications.