Retrieving Categorical Emotions Using a Probabilistic Framework to Define Preference Learning Samples

@inproceedings{Lotfian2016RetrievingCE,
  title={Retrieving Categorical Emotions Using a Probabilistic Framework to Define Preference Learning Samples},
  author={Reza Lotfian and Carlos Busso},
  booktitle={INTERSPEECH},
  year={2016}
}
Preference learning is an appealing approach for affective recognition. Instead of predicting the underlying emotional class of a sample, this framework relies on pairwise comparisons to rank-order the testing data according to an emotional dimension. This framework is relevant not only for continuous attributes such as arousal or valence, but also for categorical classes (e.g., is this sample happier than the other?). A preference learning system for categorical classes can have applications… 

Figures and Tables from this paper

Using Agreement on Direction of Change to Build Rank-Based Emotion Classifiers
TLDR
This study explores the use of relative labels to train machine learning algorithms that can rank expressive behaviors by relying on the qualitative agreement (QA) analysis to estimate relative labels which are used to train rank-based classifiers (rankers).
Formulating emotion perception as a probabilistic model with application to categorical emotion classification
  • Reza Lotfian, C. Busso
  • Computer Science
    2017 Seventh International Conference on Affective Computing and Intelligent Interaction (ACII)
  • 2017
TLDR
This study proposes a new formulation, where the emotional perception of a stimuli is a multidimensional Gaussian random variable with an unobserved distribution, and the covariance matrix of this distribution captures the intrinsic dependencies between different emotional categories.
The ordinal nature of emotions
TLDR
The thesis that emotions are by nature relative is supported by both theoretical arguments and evidence, and opens new horizons for the way emotions are viewed, represented and analyzed computationally.
Predicting Categorical Emotions by Jointly Learning Primary and Secondary Emotions through Multitask Learning
TLDR
This work takes advantage of both types of annotations to improve the performance of emotion classification and shows that considering secondary emotion labels during the learning process leads to relative improvements of 7.9% in F1-score for an 8-class emotion classification task.
Preference-Learning with Qualitative Agreement for Sentence Level Emotional Annotations
TLDR
The experimental evaluation shows that preference-learning methods to rank-order emotional attributes trained with the proposed QAbased labels achieve significantly better performance than the same algorithms trained with relative scores obtained by averaging absolute scores across annotators.
The Ordinal Nature of Emotions: An Emerging Approach
TLDR
The thesis that emotions are by nature ordinal is supported by both theoretical arguments and evidence, and opens new horizons for the way emotions are viewed, represented and analyzed computationally.
Quantifying Emotional Similarity in Speech
TLDR
This study proposes the novel formulation of measuring emotional similarity between speech recordings, and creates a representation using a DNN trained with the triplet loss function, which relies on triplets formed with an anchor, a positive example, and a negative example.
Development of Emotion Rankers Based on Intended and Perceived Emotion Labels
TLDR
This paper proposes a novel method to derive relative labels for preference learning using both the intended labels during emotion expression and the perceived labels given by all raters during perceptual evaluation, and proposes three pairwise ranking rules to generate multi-scale relevant scores for preference learn.
Ranking emotional attributes with deep neural networks
TLDR
This study uses a deep learning ranker implemented with the RankNet algorithm to evaluate preference between emotional sentences in terms of dimensional attributes (arousal, valence and dominance) and shows improved performance over ranking algorithms trained with support vector machine (SVM).
Over-Sampling Emotional Speech Data Based on Subjective Evaluations Provided by Multiple Individuals
TLDR
It is argued that several labels provided by different individuals convey more information than the consensus labels, which can help in building more robust classifiers which maximize the utilization of labeled data.
...
1
2
...

References

SHOWING 1-10 OF 33 REFERENCES
Practical considerations on the use of preference learning for ranking emotional speech
  • Reza Lotfian, C. Busso
  • Computer Science
    2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2016
TLDR
Practical considerations in training ranking classifiers including optimum number of pairs used during training, and the margin used to define the relative labels are analyzed, which are important to improve the performance of the emotional retrieval system.
Speaker-sensitive emotion recognition via ranking: Studies on acted and spontaneous speech
Don’t Classify Ratings of Affect; Rank Them!
TLDR
The main results of the paper suggest that ratings (when used) should be naturally transformed to ordinal (ranked) representations for obtaining more reliable and generalisable models of affect.
A Regression Approach to Music Emotion Recognition
TLDR
This paper forms MER as a regression problem to predict the arousal and valence values (AV values) of each music sample directly and applies the regression approach to detect the emotion variation within a music selection and find the prediction accuracy superior to existing works.
Feeler : Emotion Classification of Text Using Vector Space Model
TLDR
An emotion enabled video player which automatically detects the emotion from subtitle text of video and displays corresponding emoticon is developed and tested and discussed the results of classification using cross-validation technique for emotion classification and sentiment analyses on both the ISEAR and SemEval datasets.
Collective indexing of emotions in images. A study in emotional information retrieval
TLDR
Green's, Rorissa’s, and Iyer's analyses do not deal with emotional concepts, so further research is necessary to adjust their findings to emotions.
Recording audio-visual emotional databases from actors : a closer look
TLDR
The main limitations of the current settings used in collecting acted emotional databases are discussed, and guidelines for the design of new corpora recorded from actors that may reduce the gap observed between the laboratory condition and real-life applications are suggested.
Queries and tags in affect-based multimedia retrieval
TLDR
Results show that the use of arousal and valence as affective tags can improve retrieval results.
MSP-IMPROV: An Acted Corpus of Dyadic Interactions to Study Emotion Perception
TLDR
The MSP-IMPROV corpus is presented, a multimodal emotional database, where the goal is to have control over lexical content and emotion while also promoting naturalness in the recordings, leveraging the large size of the audiovisual database.
The SEMAINE Database: Annotated Multimodal Records of Emotionally Colored Conversations between a Person and a Limited Agent
TLDR
A large audiovisual database is created as a part of an iterative approach to building Sensitive Artificial Listener agents that can engage a person in a sustained, emotionally colored conversation.
...
1
2
3
4
...