Verbal and non-verbal cues in the communication of emotions

Polzin, Thomas

doi:10.1109/icassp.2000.859332

Cited by 22 publications

(14 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In general, neutral speech for each speaker may not be available. In that scenario, at least gender normalization should be applied [42]. Our ultimate goal is to design a framework to automatically detect emotional speech regions from large amounts of data in an unsupervised manner (e.g., call center data).…”

Section: Discussionmentioning

confidence: 99%

Analysis of Emotionally Salient Aspects of Fundamental Frequency for Emotion Detection

Busso

Lee

Narayanan

2009

IEEE Trans. Audio Speech Lang. Process.

235

View full text Add to dashboard Cite

Abstract-During expressive speech, the voice is enriched to convey not only the intended semantic message but also the emotional state of the speaker. The pitch contour is one of the important properties of speech that is affected by this emotional modulation. Although pitch features have been commonly used to recognize emotions, it is not clear what aspects of the pitch contour are the most emotionally salient. This paper presents an analysis of the statistics derived from the pitch contour. First, pitch features derived from emotional speech samples are compared with the ones derived from neutral speech, by using symmetric Kullback-Leibler distance. Then, the emotionally discriminative power of the pitch features is quantified by comparing nested logistic regression models. The results indicate that gross pitch contour statistics such as mean, maximum, minimum, and range are more emotionally prominent than features describing the pitch shape. Also, analyzing the pitch statistics at the utterance level is found to be more accurate and robust than analyzing the pitch statistics for shorter speech regions (e.g., voiced segments). Finally, the best features are selected to build a binary emotion detection system for distinguishing between emotional versus neutral speech. A new two-step approach is proposed. In the first step, reference models for the pitch features are trained with neutral speech, and the input features are contrasted with the neutral model. In the second step, a fitness measure is used to assess whether the input speech is similar to, in the case of neutral speech, or different from, in the case of emotional speech, the reference models. The proposed approach is tested with four acted emotional databases spanning different emotional categories, recording settings, speakers and languages. The results show that the recognition accuracy of the system is over 77% just with the pitch features (baseline 50%). When compared to conventional classification schemes, the proposed approach performs better in terms of both accuracy and robustness.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of Emotionally Salient Aspects of Fundamental Frequency for Emotion Detection

Busso

Lee

Narayanan

2009

IEEE Trans. Audio Speech Lang. Process.

235

View full text Add to dashboard Cite

show abstract

“…In general, people can recognize an emotional expression in neutral-content speech with about 60 percent accuracy, choosing from among about six different affective labels [10]. Computer algorithms match or slightly beat this accuracy, e.g., [11], [12]. Note that computer speech recognition that works at about 90 percent accuracy on neutrally-spoken speech tends to drop to 50-60 percent accuracy on emotional speech [13].…”

Section: Related Researchmentioning

confidence: 99%

Toward machine emotional intelligence: analysis of affective physiological state

Picard

Vyzas²,

Healey

2001

IEEE Trans. Pattern Anal. Machine Intell.

1,867

1,041

View full text Add to dashboard Cite

AbstractÐThe ability to recognize emotion is one of the hallmarks of emotional intelligence, an aspect of human intelligence that has been argued to be even more important than mathematical and verbal intelligences. This paper proposes that machine intelligence needs to include emotional intelligence and demonstrates results toward this goal: developing a machine's ability to recognize human affective state given four physiological signals. We describe difficult issues unique to obtaining reliable affective data and collect a large set of data from a subject trying to elicit and experience each of eight emotional states, daily, over multiple weeks. This paper presents and compares multiple algorithms for feature-based recognition of emotional state from this data. We analyze four physiological signals that exhibit problematic day-to-day variations: The features of different emotions on the same day tend to cluster more tightly than do the features of the same emotion on different days. To handle the daily variations, we propose new features and algorithms and compare their performance. We find that the technique of seeding a Fisher Projection with the results of Sequential Floating Forward Search improves the performance of the Fisher Projection and provides the highest recognition rates reported to date for classification of affect from physiology: 81 percent recognition accuracy on eight classes of emotion, including neutral.

show abstract

“…Provided that a person utters what he or she feels without hesitation in a controlled therapy setting, a very valuable piece of information is obtained that can be used for estimating the emotional state of a person through semantic analysis (e.g. keywords and n-grams) [12,13]. The emotional state estimator is part of an ongoing research project: Adaptive control of virtual reality scenarios in therapy of posttraumatic stress disorder (PTSD), at University of Zagreb [14,15].…”

Section: Introductionmentioning

confidence: 99%

Development of Acoustic Model for Croatian Language Using HTK

Dropuljić¹,

Petrinovic

2010

Automatika

View full text Add to dashboard Cite

Original scientific paperPaper presents development of the acoustic model for Croatian language for automatic speech recognition (ASR). Continuous speech recognition is performed by means of the Hidden Markov Models (HMM) implemented in the HMM Toolkit (HTK). In order to adjust the HTK to the native language a novel algorithm for Croatian language transcription (CLT) has been developed. It is based on phonetic assimilation rules that are applied within uttered words. Phonetic questions for state tying of different triphone models have also been developed. The automated system for training and evaluation of acoustic models has been developed and integrated with the new graphical user interface (GUI). Targeted applications of this ASR system are stress inoculation training (SIT) and virtual reality exposure therapy (VRET). Adaptability of the model to a closed set of speakers is important for such applications and this paper investigates the applicability of the HTK tool for typical scenarios. Robustness of the tool to a new language was tested in matched conditions by a parallel training of an English model that was used as a baseline. Ten native Croatian speakers participated in experiments. Encouraging results were achieved and reported with the developed model for Croatian language.Key words: Acoustic model, Automatic speech recognition, Croatian language, Hidden Markov models, Phonetic assimilation, Phonetic transcription algorithm, Recognition accuracy Razvoj akustičkog modela hrvatskog jezika pomoću alata HTK. Rad opisuje razvoj akustičkog modela hrvatskog jezika za potrebe sustava za automatsko prepoznavanje govora. Prepoznavanje prirodnog spojenog izgovora ostvaruje se korištenjem skrivenih Markovljevih modela (HMM) u okviru alata HTK. U svrhu prilagodbe ovog alata na hrvatski jezik razvijen je novi algoritam za automatsku fonetsku transkripciju hrvatskih riječi. Zasniva se na načelu fonetske asimilacije unutar izgovorenih riječi. Razvijen je i skup fonetskih pitanja koji se koristi za klasifikaciju prilikom udruživanja trifonskih modela sličnih glasova. Razvijena je automatizirana aplikacija za gradnju i evaluaciju akustičkih modela, integrirana s novo razvijenim grafičkim sučeljem. Primjene ovog sustava za prepoznavanje su trening s doziranim izlaganjem stresu (SIT) i terapija izlaganjem primjenom virtualne stvarnosti (VRET). Prilagodljivost akustičkog modela na zatvoren skup govornika vrlo je važna za takve primjene, pa se u radu istražuje primjenjivost alata HTK u tipičnim scenarijima. Robusnost alata na promjenu jezika istražuje se uparenim treniranjem i evaluacijom ekvivalentnog modela engleskog jezika u jednakim uvjetima. U eksperimentima je sudjelovalo deset izvornih hrvatskih govornika. Ostvareni rezultati za hrvatski jezik prikazani u radu pokazuju zadovoljavajuća svojstva razvijenog akustičkog modela hrvatskog jezika.Ključne riječi: akustički model, automatsko prepoznavanje govora, hrvatski jezik, skriveni Markovljevi modeli, algoritam za fonetsku transkripciju, fonetska asimilacija, točnost pr...

show abstract

Verbal and non-verbal cues in the communication of emotions

Cited by 22 publications

References 8 publications

Analysis of Emotionally Salient Aspects of Fundamental Frequency for Emotion Detection

Analysis of Emotionally Salient Aspects of Fundamental Frequency for Emotion Detection

Toward machine emotional intelligence: analysis of affective physiological state

Development of Acoustic Model for Croatian Language Using HTK

Contact Info

Product

Resources

About