Deep Visual Attributes vs. Hand-Crafted Audio Features on Multidomain Speech Emotion Recognition

Papakostas, Michalis; Spyrou, Evaggelos; Γιαννακόπουλος, Θεόδωρος; Siantikos, Giorgos; Sgouropoulos, Dimitrios; Mylonas, Phivos; Makedon, Fillia

doi:10.3390/computation5020026

Cited by 45 publications

(29 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Investigating the content of past customer-firm interactions rather than simply focusing on the current interaction could capture changes in the relationship and help identify the customers who are most likely to churn. Deep learning methods have also been developed to analyze emotions based on audio content (e.g., Papakostas et al 2017), which can be used to assess a customer's mindset across a sequence of interactions so as to spot changes in how customers communicate with the firm that may signal changes in the underlying relationship.…”

Section: Challenges To Leveraging Unstructured Customer-firm Interactmentioning

confidence: 99%

Capturing Marketing Information to Fuel Growth

Du¹,

Netzer²,

Schweidel³

et al. 2020

Journal of Marketing

View full text Add to dashboard Cite

Marketing is the functional area primarily responsible for driving the organic growth of a firm. In the age of digital marketing and big data, marketers are inundated with increasingly rich data from an ever-expanding array of sources. Such data may help marketers generate insights about customers and competitors. One fundamental question remains: How can marketers wrestle massive flows of existing and nascent data resources into coherent, effective growth strategies? Against such a backdrop, the Marketing Science Institute has made “capturing information to fuel growth” a top research priority. We begin by discussing the streetlight effect—an overreliance on readily available data due to ease of measurement and application—as contributing to the disconnect between marketing data growth and firm growth. We then use the customer equity framework to structure our discussion of six areas where we see substantial under-tapped opportunities—incorporating social network and biometric data in customer acquisition, trend and competitive interaction data in customer development, and unstructured and causal data in customer retention. We highlight challenges that obstruct firms from realizing such data-driven growth opportunities and how future research may help overcome those challenges.

show abstract

Section: Challenges To Leveraging Unstructured Customer-firm Interactmentioning

confidence: 99%

Capturing Marketing Information to Fuel Growth

Du¹,

Netzer²,

Schweidel³

et al. 2020

Journal of Marketing

View full text Add to dashboard Cite

show abstract

“…Our study also used raw spectrogram as the input data to our proposed model. We therefore extracted spectrogram from all the emotional audio signal in each corpora by cropping segment of 2 s length randomly using the short-time Fourier transforms at a window size short term of 40 ms and steps of 20 ms proposed by Reference [65].…”

Section: Emotion Corporamentioning

confidence: 99%

Cross lingual speech emotion recognition via triple attentive asymmetric convolutional neural network

et al. 2020

View full text Add to dashboard Cite

The application of cross-corpus for speech emotion recognition (SER) via domain adaptation methods have gain high acknowledgment for developing good robust emotion recognition systems using different corpora or datasets. However, the issue of crosslingual still remains a challenge in SER and needs more attention to resolve the scenario of applying different language types in both training and testing. In this paper, we propose a triple attentive asymmetric convolutional neural network to address the recognition of emotions for cross-lingual and crosscorpus speech in an unsupervised approach. The proposed method adopts the joint supervision of softmax loss and center loss to learn high power discriminative feature representations for target domain via the use of high quality pseudo-labels. The proposed model uses three attentive convolutional neural networks asymmetrically, where two of the networks are used to artificially label unlabeled target samples as a result of their predictions from training on source labeled samples and the other network is used to obtain salient target discriminative features from the pseudo-labeled target samples. We evaluate our proposed method on three different language types (i.e., English, German, and Italian) data sets. The experimental results indicate that, our proposed method achieves higher prediction accuracy over other state-of-the-art methods.

show abstract

“…[9,10], or in other cases [14][15][16] follow the dimensional approach [17] that originates from psychophysiology. Spectrograms have also been previously used for other audio analysis-related tasks, such as content classification [18] and segmentation [19], for stress recognition [20] and more recently for emotion recognition with convolutional neural networks [21,22].…”

Section: Non-linguistic Approaches For Emotion Recognitionmentioning

confidence: 99%

Emotion Recognition from Speech Using the Bag-of-Visual Words on Audio Segment Spectrograms

et al. 2019

Self Cite

View full text Add to dashboard Cite

It is noteworthy nowadays that monitoring and understanding a human’s emotional state plays a key role in the current and forthcoming computational technologies. On the other hand, this monitoring and analysis should be as unobtrusive as possible, since in our era the digital world has been smoothly adopted in everyday life activities. In this framework and within the domain of assessing humans’ affective state during their educational training, the most popular way to go is to use sensory equipment that would allow their observing without involving any kind of direct contact. Thus, in this work, we focus on human emotion recognition from audio stimuli (i.e., human speech) using a novel approach based on a computer vision inspired methodology, namely the bag-of-visual words method, applied on several audio segment spectrograms. The latter are considered to be the visual representation of the considered audio segment and may be analyzed by exploiting well-known traditional computer vision techniques, such as construction of a visual vocabulary, extraction of speeded-up robust features (SURF) features, quantization into a set of visual words, and image histogram construction. As a last step, support vector machines (SVM) classifiers are trained based on the aforementioned information. Finally, to further generalize the herein proposed approach, we utilize publicly available datasets from several human languages to perform cross-language experiments, both in terms of actor-created and real-life ones.

show abstract

Deep Visual Attributes vs. Hand-Crafted Audio Features on Multidomain Speech Emotion Recognition

Cited by 45 publications

References 38 publications

Capturing Marketing Information to Fuel Growth

Capturing Marketing Information to Fuel Growth

Cross lingual speech emotion recognition via triple attentive asymmetric convolutional neural network

Emotion Recognition from Speech Using the Bag-of-Visual Words on Audio Segment Spectrograms

Contact Info

Product

Resources

About