Decision level combination of multiple modalities for recognition and analysis of emotional expression

Metallinou, Angeliki; Lee, Sungbok; Narayanan, Shrikanth

doi:10.1109/icassp.2010.5494890

Cited by 77 publications

(57 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results show that EUs contain different levels of emotion salience, demonstrating that certain EUs are more strongly associated with particular emotion classes (e.g., a rise in anger strongly suggests the class of "angry"). The classification results demonstrate that this method can achieve comparable results to the state of the art [15]. These findings suggest that there may exist basic building blocks that underlie expressions of emotion.…”

Section: Introductionsupporting

confidence: 64%

See 1 more Smart Citation

Identifying salient sub-utterance emotion dynamics using flexible units and estimates of affective flow

Provost

2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

Emotion recognition is the process of identifying the affective characteristics of an utterance given either static or dynamic descriptions of its signal content. This requires the use of units, windows over which the emotion variation is quantified. However, the appropriate time scale for these units is still an open question. Traditionally, emotion recognition systems have relied upon units of fixed length, whose variation is then modeled over time. This paper takes the view that emotion is expressed over units of variable length. In this paper, variable-length units are introduced and used to capture the local dynamics of emotion at the sub-utterance scale. The results demonstrate that subsets of these local dynamics are salient with respect to emotion class. These salient units provide insight into the natural variation in emotional speech and can be used in a classification framework to achieve performance comparable to the state-of-theart. This hints at the existence of building blocks that may underlie natural human emotional communication.

show abstract

Section: Introductionsupporting

confidence: 64%

“…The results demonstrate that variable-length units can effectively capture local dynamics and can be used in a classification framework to achieve results comparable to the state-of-the-art on the IEMOCAP database (62.42% [15]). The results also show that these units can be used to interpret the emotional dynamics of affective utterances.…”

Section: Discussionmentioning

confidence: 62%

Identifying salient sub-utterance emotion dynamics using flexible units and estimates of affective flow

Provost

2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

show abstract

“…We also compare our results with the maximal accuracy achieved from a previous work of Metallinou et al [37], which utilizes the same IEMOCAP database as our work and introduces a decisionlevel Bayesian fusion over models using face, voice, and head movement cues. Although Metallinou's work used a different subset of the IEMOCAP database, this comparison supports the strong performance of our proposed method.…”

Section: Baseline Modelsmentioning

confidence: 50%

Deep learning for robust feature generation in audiovisual emotion recognition

Kim

Lee

Provost

2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

313

181

View full text Add to dashboard Cite

Automatic emotion recognition systems predict high-level affective content from low-level human-centered signal cues. These systems have seen great improvements in classification accuracy, due in part to advances in feature selection methods. However, many of these feature selection methods capture only linear relationships between features or alternatively require the use of labeled data. In this paper we focus on deep learning techniques, which can overcome these limitations by explicitly capturing complex non-linear feature interactions in multimodal data. We propose and evaluate a suite of Deep Belief Network models, and demonstrate that these models show improvement in emotion classification performance over baselines that do not employ deep learning. This suggests that the learned highorder non-linear relationships are effective for emotion recognition.

show abstract

“…This way of combining evidence lead to overall improvements in our early work. However much work on ensemble learning has demonstrated that for variety of tasks this method of combination is not as powerful as decision-level combination (for example see Raaijmakers, Truong, & Wilson, 2008;van Halteren, Zavrel, & Daelemans, 1998;Metallinou, Lee, & Narayanan, 2010;Bertolami & Bunke, 2006). We treat the feature-level combination as the baseline for our experiments.…”

Section: Feature-level Combination(c1)mentioning

confidence: 99%

Combining Lexical and Syntactic Features for Detecting Content-Dense Texts in News

Yang¹,

Nenkova²

2017

jair

View full text Add to dashboard Cite

Content-dense news report important factual information about an event in direct, succinct manner. Information seeking applications such as information extraction, question answering and summarization normally assume all text they deal with is content-dense. Here we empirically test this assumption on news articles from the business, U.S. international relations, sports and science journalism domains. Our findings clearly indicate that about half of the news texts in our study are in fact not content-dense and motivate the development of a supervised content-density detector. We heuristically label a large training corpus for the task and train a two-layer classifying model based on lexical and unlexicalized syntactic features. On manually annotated data, we compare the performance of domain-specific classifiers, trained on data only from a given news domain and a general classifier in which data from all four domains is pooled together. Our annotation and prediction experiments demonstrate that the concept of content density varies depending on the domain and that naive annotators provide judgement biased toward the stereotypical domain label. Domain-specific classifiers are more accurate for domains in which contentdense texts are typically fewer. Domain independent classifiers reproduce better naive crowdsourced judgements. Classification prediction is high across all conditions, around 80%.

show abstract

Decision level combination of multiple modalities for recognition and analysis of emotional expression

Cited by 77 publications

References 12 publications

Identifying salient sub-utterance emotion dynamics using flexible units and estimates of affective flow

Identifying salient sub-utterance emotion dynamics using flexible units and estimates of affective flow

Deep learning for robust feature generation in audiovisual emotion recognition

Combining Lexical and Syntactic Features for Detecting Content-Dense Texts in News

Contact Info

Product

Resources

About