The rise of machine-learning systems that process sensory input has brought with it a rise in comparisons between human and machine perception. But such comparisons face a challenge: Whereas machine perception of some stimulus can often be probed through direct and explicit measures, much of human perceptual knowledge is latent, incomplete, or unavailable for explicit report. Here, we explore how this asymmetry can cause such comparisons to misestimate the overlap in human and machine perception. As a case study, we consider human perception of \textit{adversarial speech} -- synthetic audio commands that are recognized as valid messages by automated speech-recognition systems but that human listeners reportedly hear as meaningless noise. In five experiments, we adapt task designs from the human psychophysics literature to show that even when subjects cannot freely transcribe such speech commands (the previous benchmark for human understanding), they often can demonstrate other forms of understanding, including discriminating adversarial speech from closely matched non-speech (Experiments 1--2), finishing common phrases begun in adversarial speech (Experiments 3--4), and solving simple math problems posed in adversarial speech (Experiment 5) -- even for stimuli previously described as unintelligible to human listeners. We recommend the adoption of such "sensitive tests" when comparing human and machine perception, and we discuss the broader consequences of such approaches for assessing the overlap between systems.
Hearing Like Humans? Sound Symbolism and Perceptual Alignment in Speech Language Models
Graduate Institute of Communication Engineering National Taiwan University Taipei, Taiwan；Graduate Institute of Electrical Engineering National Taiwan University Taipei, Taiwan；Artificial Intelligence Center of Research Excellence National Taiwan University Taipei, Taiwan
The electroencephalography (EEG) signals recorded in parallel with speech are used to perform isolated and continuous speech recognition. During speaking process, one also hears his or her own speech and this speech perception is also reflected in the recorded EEG signals. In this paper we investigate whether it is possible to separate out this speech perception component from EEG signals in order to design more robust EEG based speech recognition systems. We further demonstrate predicting EEG signals recorded in parallel with speaking from EEG signals recorded in parallel with passive listeni…
Universität Oldenburg（奥尔登堡大学）；Cluster of Excellence Hearing4all（卓越集群Hearing4all）；Medizinische Physik（医学物理）
The effect of hearing impairment on speech perception was described by Plomp (1978) as a sum of a loss of class A, due to signal attenuation, and a loss of class D, due to signal distortion. While a loss of class A can be compensated by linear amplification, a loss of class D, which severely limits the benefit of hearing aids in noisy listening conditions, cannot. Not few users of hearing aids keep complaining about the limited benefit of their devices in noisy environments. Recently, in an approach to model human speech recognition by means of a re-purposed automatic speech recognition system…
Speech is a common and effective way of communication between humans, and modern consumer devices such as smartphones and home hubs are equipped with deep learning based accurate automatic speech recognition to enable natural interaction between humans and machines. Recently, researchers have demonstrated powerful attacks against machine learning models that can fool them to produceincorrect results. However, nearly all previous research in adversarial attacks has focused on image recognition and object detection models. In this short paper, we present a first of its kind demonstration of adve…
Automatic speech recognition and voice identification systems are being deployed in a wide array of applications, from providing control mechanisms to devices lacking traditional interfaces, to the automatic transcription of conversations and authentication of users. Many of these applications have significant security and privacy considerations. We develop attacks that force mistranscription and misidentification in state of the art systems, with minimal impact on human comprehension. Processing pipelines for modern systems are comprised of signal preprocessing and feature extraction steps, w…
University of Texas at Dallas（德克萨斯大学达拉斯分校）；Southern University of Science and Technology（南方科技大学）
Electrical hearing by cochlear implants (CIs) may be fundamentally different from acoustic hearing by normal-hearing (NH) listeners, presumably showing unequal speech quality perception in various noise environments. Noise reduction (NR) algorithms used in CI reduce the noise in favor of signal-to-noise ratio (SNR), regardless of plausible accompanying distortions that may degrade the speech quality perception. To gain better understanding of CI speech quality perception, the present work aimed investigating speech quality perception in a diverse noise conditions, including factors of speech/n…
Human perception of surrounding events is strongly dependent on audio cues. Thus, acoustic insulation can seriously impact situational awareness. We present an exploratory study in the domain of assistive computing, eliciting requirements and presenting solutions to problems found in the development of an environmental sound recognition system, which aims to assist deaf and hard of hearing people in the perception of sounds. To take advantage of smartphones computational ubiquity, we propose a system that executes all processing on the device itself, from audio features extraction to recogniti…
Many hearing-impaired listeners struggle to localize sounds due to poor availability of binaural cues. Listeners with a cochlear implant and a contralateral hearing aid -- so-called bimodal listeners -- are amongst the worst performers, as both interaural time and level differences are poorly transmitted. We present a new method to enhance head shadow in the low frequencies. Head shadow enhancement is achieved with a fixed beamformer with contralateral attenuation in each ear. The method results in interaural level differences which vary monotonically with angle. It also improves low-frequency…
Do You Hear What I Mean? Quantifying the Instruction-Perception Gap in Instruction-Guided Expressive Text-To-Speech Systems
National Taiwan University（台湾大学）；University of Southern California（南加州大学）；University of Michigan（密歇根大学）；NTU Artificial Intelligence Center of Research Excellence (NTU AI-CoRE)（台湾大学人工智能卓越研究中心）
Hörzentrum Oldenburg gGmbH（奥尔登堡听力中心有限责任公司）；University of Oldenburg（奥尔登堡大学）
Objective: Distorted loudness perception is one of the main complaints of hearing aid users. Being able to measure loudness perception correctly in the clinic is essential for fitting hearing aids. For this, experiments in the clinic should be able to reflect and capture loudness perception as in everyday-life situations. Little research has been done comparing loudness perception in the field and in the laboratory. Design: Participants rated the loudness in the field and in the laboratory of 36 driving actions done by four different vehicles. The field measurements were done in a restricted s…
In the Clarity project, we will run a series of machine learning challenges to revolutionise speech processing for hearing devices. Over five years, there will be three paired challenges. Each pair will consist of a competition focussed on hearing-device processing ("enhancement") and another focussed on speech perception modelling ("prediction"). The enhancement challenges will deliver new and improved approaches for hearing device signal processing for speech. The parallel prediction challenges will develop and improve methods for predicting speech intelligibility and quality for hearing imp…
In daily life, social interaction and acoustic communication often take place in complex acoustic environments (CAE) with a variety of interfering sounds and reverberation. For hearing research and the evaluation of hearing systems, simulated CAEs using virtual reality techniques have gained interest in the context of ecological validity. In the current study, the effect of scene complexity and visual representation of the scene on psychoacoustic measures like sound source location, distance perception, loudness, speech intelligibility, and listening effort in a virtual audio-visual environmen…
From Spikes to Speech: NeuroVoc -- A Biologically Plausible Vocoder Framework for Auditory Perception and Cochlear Implant Simulation
Leiden Institute of Advanced Computer Science（莱顿高级计算机科学研究所）；Department of Otorhinolaryngology（耳鼻喉科部门）；Leiden Institute for Brain and Cognition（莱顿脑认知研究所）；Bioelectronics group, EEMCS（生物电子组，EEMCS）
Eindhoven University of Technology（埃因霍温理工大学）；Ghent University（根特大学）
In this paper the auditory model developed by Dau et al. [J. Acoust. Soc. Am. 102, 2892-2905 (1997)] was used to simulate the perceptual similarity between complex sounds. For this purpose, a central processor stage was developed and attached as a back-end module to the auditory model. As complex sounds, a set of recordings of one note played on seven different pianos was used, whose similarity has been recently measured by Osses et al. [J. Acoust. Soc. Am. 146, 1024-1035 (2019)] using a 3-AFC discrimination task in noise. The auditory model has several processing stages that are to a greater …
Evaluating Speech-in-Speech Perception via a Humanoid Robot
University Medical Center Groningen（格罗宁根大学医学中心）；University of Groningen（格罗宁根大学）；Pento Audiology Centre（Pento听力中心）；Lyon Neuroscience Research Center（里昂神经科学研究中心）；Université Claude Bernard Lyon 1（里昂第一大学）；Université de Lyon（里昂大学）
