Improving speaker diarization for naturalistic child-adult conversational interactions using contextual information.
Ontology highlight
ABSTRACT:
SUBMITTER: Kumar M
PROVIDER: S-EPMC7030978 | biostudies-literature | 2020 Feb
REPOSITORIES: biostudies-literature
Publications
Improving speaker diarization for naturalistic child-adult conversational interactions using contextual information.
Kumar Manoj M Kim So Hyun SH Lord Catherine C Narayanan Shrikanth S
The Journal of the Acoustical Society of America 20200201 2
While deep learning has driven recent improvements in audio speaker diarization, it often faces performance issues in challenging interaction scenarios and varied acoustic settings such as between a child and adult (caregiver/examiner). In this work, the role of contextual factors that affect diarization performance in such interactions is analyzed. Factors that affect each type of diarization error are identified. Furthermore, a DNN is trained on diarization outputs in conjunction with the fact ...[more]
Similar Datasets
Project description:When infants and adults communicate, they exchange social signals of availability and communicative intention such as eye gaze. Previous research indicates that when communication is successful, close temporal dependencies arise between adult speakers' and listeners' neural activity. However, it is not known whether similar neural contingencies exist within adult-infant dyads. Here, we used dual-electroencephalography to assess whether direct gaze increases neural coupling between adults and infants during screen-based and live interactions. In experiment 1 (n = 17), infants viewed videos of an adult who was singing nursery rhymes with (i) direct gaze (looking forward), (ii) indirect gaze (head and eyes averted by 20°), or (iii) direct-oblique gaze (head averted but eyes orientated forward). In experiment 2 (n = 19), infants viewed the same adult in a live context, singing with direct or indirect gaze. Gaze-related changes in adult-infant neural network connectivity were measured using partial directed coherence. Across both experiments, the adult had a significant (Granger) causal influence on infants' neural activity, which was stronger during direct and direct-oblique gaze relative to indirect gaze. During live interactions, infants also influenced the adult more during direct than indirect gaze. Further, infants vocalized more frequently during live direct gaze, and individual infants who vocalized longer also elicited stronger synchronization from the adult. These results demonstrate that direct gaze strengthens bidirectional adult-infant neural connectivity during communication. Thus, ostensive social signals could act to bring brains into mutual temporal alignment, creating a joint-networked state that is structured to facilitate information transfer during early communication and learning.
| S-EPMC5740679 | biostudies-literature
Project description:Human language is extremely versatile, combining a limited set of signals in an unlimited number of ways. However, it is unknown whether conversational visual signals feed into the composite utterances with which speakers communicate their intentions. We assessed whether different combinations of visual signals lead to different intent interpretations of the same spoken utterance. Participants viewed a virtual avatar uttering spoken questions while producing single visual signals (i.e., head turn, head tilt, eyebrow raise) or combinations of these signals. After each video, participants classified the communicative intention behind the question. We found that composite utterances combining several visual signals conveyed different meaning compared to utterances accompanied by the single visual signals. However, responses to combinations of signals were more similar to the responses to related, rather than unrelated, individual signals, indicating a consistent influence of the individual visual signals on the whole. This study therefore provides first evidence for compositional, non-additive (i.e., Gestalt-like) perception of multimodal language.
| S-EPMC10821935 | biostudies-literature
Project description:BackgroundDue to time limitations, the preanesthetic consultation (PAC) is not the best time for patients to integrate information specific to their perioperative care pathway.ObjectiveThe main objectives of this study were to evaluate the effectiveness of a digital companion on patients' knowledge of anesthesia and their satisfaction after real-life implementation.MethodsWe conducted a prospective, monocentric, comparative study using a before-and-after design. In phase 1, a 9-item self-reported anesthesia knowledge test (Delphi method) was administered to patients before and after their PAC (control group: PAC group). In phase 2, the study was repeated immediately after the implementation of a digital conversational agent, MyAnesth (@+PAC group). Patients' satisfaction and their representations for anesthesia were also assessed using a Likert scale and the Abric method of hierarchized evocation.ResultsA total of 600 tests were distributed; 205 patients and 98 patients were included in the PAC group and @+PAC group, respectively. Demographic characteristics and mean scores on the 9-point preinformation test (PAC group: 4.2 points, 95% CI 3.9-4.4; @+PAC: 4.3 points, 95% CI 4-4.7; P=.37) were similar in the two groups. The mean score after receiving information was better in the @+PAC group than in the PAC group (6.1 points, 95% CI 5.8-6.4 points versus 5.2 points, 95% CI 5.0-5.4 points, respectively; P<.001), with an added value of 0.7 points (95% CI 0.3-1.1; P<.001). Among the respondents in the @+PAC group, 82% found the information to be clear and appropriate, and 74% found it easily accessible. Before receiving information, the central core of patients' representations for anesthesia was focused on the fear of being put to sleep and thereafter on caregiver skills and comfort.ConclusionsThe implementation of our digital conversational agent in addition to the PAC improved patients' knowledge about their perioperative care pathway. This innovative audiovisual support seemed clear, adapted, easily accessible, and reassuring. Future studies should focus on adapting both the content and delivery of a digital conversational agent for the PAC in order to maximize its benefit to patients.
| S-EPMC7748965 | biostudies-literature
Project description:Recent research has extensively reported the phenomenon of inter-brain neural coupling between speaker and listener during speech communication. Yet, the specific speech processes underlying this neural coupling remain elusive. To bridge this gap, this study estimated the correlation between the temporal dynamics of speaker-listener neural coupling with speech features, utilizing two inter-brain datasets accounting for different noise levels and listener's language experiences (native vs. non-native). We first derived time-varying speaker-listener neural coupling, extracted acoustic feature (envelope) and semantic features (entropy and surprisal) from speech, and then explored their correlational relationship. Our findings reveal that in clear conditions, speaker-listener neural coupling correlates with semantic features. However, as noise increases, this correlation is only significant for native listeners. For non-native listeners, neural coupling correlates predominantly with acoustic feature rather than semantic features. These results revealed how speaker-listener neural coupling associated with the acoustic and semantic features under various scenarios, enriching our understanding of the inter-brain neural mechanisms during natural speech communication. We therefore advocate for more attention on the dynamic nature of speaker-listener neural coupling and its modelling with multi-level speech features.
| S-EPMC11296674 | biostudies-literature
Project description:Episodic memories are contextual experiences ordered in time. This is underpinned by associative binding between events within the same contexts. The role of prediction errors in declarative memory is well established but has not been investigated in the time dimension of complex episodic memories. Here we combine these two properties of episodic memory, extend them into the temporal domain and demonstrate that prediction errors in different naturalistic contexts lead to changes in the temporal ordering of event structures in them. The wrongly predicted older sequences were weakened despite their reactivation. Interestingly the newly encoded sequences with prediction errors, seen once, showed accuracy as high as control sequences which were viewed repeatedly without change. Drift-diffusion modelling revealed a lower decision threshold for the newer sequences than older sequences, reflected by their faster recall. Moreover, participants' adjustments to their decision threshold significantly correlated with their relative speed of sequence memory recall. These results suggest a temporally distinct and adaptive role for prediction errors in learning and reorganizing episodic temporal sequences.
| S-EPMC8196002 | biostudies-literature
Project description:Language occurs naturally in conversations. However, the study of the neural underpinnings of language has mainly taken place in single individuals using controlled language material. The interactive elements of a conversation (e.g., turn-taking) are often not part of neurolinguistic setups. The prime reason is the difficulty to combine open unrestricted conversations with the requirements of neuroimaging. It is necessary to find a trade-off between the naturalness of a conversation and the restrictions imposed by neuroscientific methods to allow for ecologically more valid studies. Here, we make an attempt to study the effects of a conversational element, namely turn-taking, on linguistic neural correlates, specifically the N400 effect. We focus on the physiological aspect of turn-taking, the speaker-switch, and its effect on the detectability of the N400 effect. The N400 event-related potential reflects expectation violations in a semantic context; the N400 effect describes the difference of the N400 amplitude between semantically expected and unexpected items. Sentences with semantically congruent and incongruent final words were presented in two turn-taking modes: (1) reading aloud first part of the sentence and listening to speaker-switch for the final word, and (2) listening to first part of the sentence and speaker-switch for the final word. A significant N400 effect was found for both turn-taking modes, which was not influenced by the mode itself. However, the mode significantly affected the P200, which was increased for the reading aloud mode compared to the listening mode. Our results show that an N400 effect can be detected during a speaker-switch. Speech articulation (reading aloud) before the analyzed sentence fragment did also not impede the N400 effect detection for the final word. The speaker-switch, however, seems to influence earlier components of the electroencephalogram, related to processing of salient stimuli. We conclude that the N400 can effectively be used to study neural correlates of language in conversational approaches including speaker-switches.
| S-EPMC5124707 | biostudies-literature
Project description:People frequently face decisions that require making inferences about withheld information. The advent of large language models coupled with conversational technology, e.g., Alexa, Siri, Cortana, and the Google Assistant, is changing the mode in which people make these inferences. We demonstrate that conversational modes of information provision, relative to traditional digital media, result in more critical responses to withheld information, including: (1) a reduction in evaluations of a product or service for which information is withheld and (2) an increased likelihood of recalling that information was withheld. These effects are robust across multiple conversational modes: a recorded phone conversation, an unfolding chat conversation, and a conversation script. We provide further evidence that these effects hold for conversations with the Google Assistant, a prominent conversational technology. The experimental results point to participants' intuitions about why the information was withheld as the driver of the effect.
| S-EPMC11008880 | biostudies-literature
Project description:Peripheral vision comprises most of our visual field, and is essential in guiding visual behavior. Its characteristic capabilities and limitations, which distinguish it from foveal vision, have been explained by the most influential theory of peripheral vision as the product of representing the visual input using summary statistics. Despite its success, this account may provide a limited understanding of peripheral vision, because it neglects processes of perceptual grouping and segmentation. To test this hypothesis, we studied how contextual modulation, namely the modulation of the perception of a stimulus by its surrounds, interacts with segmentation in human peripheral vision. We used naturalistic textures, which are directly related to summary-statistics representations. We show that segmentation cues affect contextual modulation, and that this is not captured by our implementation of the summary-statistics model. We then characterize the effects of different texture statistics on contextual modulation, providing guidance for extending the model, as well as for probing neural mechanisms of peripheral vision.
| S-EPMC7794279 | biostudies-literature
Project description:Focusing attention on one speaker on the background of other irrelevant speech can be a challenging feat. A longstanding question in attention research is whether and how frequently individuals shift their attention towards task-irrelevant speech, arguably leading to occasional detection of words in a so-called unattended message. However, this has been difficult to gauge empirically, particularly when participants attend to continuous natural speech, due to the lack of appropriate metrics for detecting shifts in internal attention. Here we introduce a new experimental platform for studying the dynamic deployment of attention among concurrent speakers, utilizing a unique combination of Virtual Reality (VR) and Eye-Tracking technology. We created a Virtual Café in which participants sit across from and attend to the narrative of a target speaker. We manipulated the number and location of distractor speakers by placing additional characters throughout the Virtual Café. By monitoring participant's eye-gaze dynamics, we studied the patterns of overt attention-shifts among concurrent speakers as well as the consequences of these shifts on speech comprehension. Our results reveal important individual differences in the gaze-pattern displayed during selective attention to speech. While some participants stayed fixated on a target speaker throughout the entire experiment, approximately 30% of participants frequently shifted their gaze toward distractor speakers or other locations in the environment, regardless of the severity of audiovisual distraction. Critically, preforming frequent gaze-shifts negatively impacted the comprehension of target speech, and participants made more mistakes when looking away from the target speaker. We also found that gaze-shifts occurred primarily during gaps in the acoustic input, suggesting that momentary reductions in acoustic masking prompt attention-shifts between competing speakers, in line with "glimpsing" theories of processing speech in noise. These results open a new window into understanding the dynamics of attention as they wax and wane over time, and the different listening patterns employed for dealing with the influx of sensory input in multisensory environments. Moreover, the novel approach developed here for tracking the locus of momentary attention in a naturalistic virtual-reality environment holds high promise for extending the study of human behavior and cognition and bridging the gap between the laboratory and real-life.
| S-EPMC6857110 | biostudies-literature
Project description:The neural mechanism for selectively tuning in to a target speaker while tuning out the others in a multi-speaker situation (i.e., the cocktail-party effect) remains elusive. Here we addressed this issue by measuring brain activity simultaneously from a listener and from multiple speakers while they were involved in naturalistic conversations. Results consistently show selectively enhanced interpersonal neural synchronization (INS) between the listener and the attended speaker at left temporal-parietal junction, compared with that between the listener and the unattended speaker across different multi-speaker situations. Moreover, INS increases significantly prior to the occurrence of verbal responses, and even when the listener's brain activity precedes that of the speaker. The INS increase is independent of brain-to-speech synchronization in both the anatomical location and frequency range. These findings suggest that INS underlies the selective process in a multi-speaker situation through neural predictions at the content level but not the sensory level of speech.
| S-EPMC6008393 | biostudies-literature