Analytical Features: A Knowledge-Based Approach to Audio Feature Generation

Pachet, François; Roy, Pierre

doi:10.1155/2009/153017

Cited by 38 publications

(26 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, we can try to make better predictions by using a larger ground truth dataset or by designing new audio descriptors especially relevant for this task. Another option would be to generate analytical features [30], or to combine several classifiers to try to increase the accuracy of the system. We could also consider the use of other contextual information like metadata, tags, or text found on the Internet (from music blogs for instance).…”

Section: Discussionmentioning

confidence: 99%

Indexing music by mood: design and integration of an automatic content-based annotator

Laurier

Meyers

Serrà

et al. 2009

Multimed Tools Appl

View full text Add to dashboard Cite

In the context of content analysis for indexing and retrieval, a method for creating automatic music mood annotation is presented. The method is based on results from psychological studies and framed into a supervised learning approach using musical features automatically extracted from the raw audio signal. We present here some of the most relevant audio features to solve this problem. A ground truth, used for training, is created using both social network information systems (wisdom of crowds) and individual experts (wisdom of the few). At the experimental level, we evaluate our approach on a database of 1000 songs. Tests of different classification methods, configurations and optimizations have been conducted, showing that Support Vector Machines perform best for the task at hand. Moreover, we evaluate the algorithm robustness against different audio compression schemes. This fact, often neglected, is fundamental to build a system that is usable in real conditions. In addition, the integration of a fast and scalable version of this technique with the European Project PHAROS is discussed. This real world application demonstrates the usability of this tool to annotate large-scale databases. We also report on a user evaluation in the context of the PHAROS search engine, asking people about the utility, interest and innovation of this technology in real world use cases.

show abstract

Section: Discussionmentioning

confidence: 99%

Indexing music by mood: design and integration of an automatic content-based annotator

Laurier

Meyers

Serrà

et al. 2009

Multimed Tools Appl

View full text Add to dashboard Cite

show abstract

“…The large number of LLD and functionals has recently promoted the extraction of very large feature vectors (brute-force extraction), up to many thousands of features obtained either by analytical feature generation [226,199,142] or, in a few studies, by evolutionary generation [185]. Such brute-forcing also often includes hierarchical functional application (e. g., mean of maxima) to better cope with statistical outliers.…”

Section: Featuresmentioning

confidence: 99%

Recognising realistic emotions and affect in speech: State of the art and lessons learnt from the first challenge

Schuller

Batliner

Steidl

et al. 2011

Speech Communication

606

329

View full text Add to dashboard Cite

More than a decade has passed since research on automatic recognition of emotion from speech has become a new field of research in line with its 'big brothers' speech and speaker recognition. This article attempts to provide a short overview on where we are today, how we got there and what this can reveal us on where to go next and how we could arrive there. In a first part, we address the basic phenomenon reflecting the last fifteen years, commenting on databases, modelling and annotation, the unit of analysis and prototypicality. We then shift to automatic processing including discussions on features, classification, robustness, evaluation, and implementation and system integration. From there we go to the first comparative challenge on emotion recognition from speech -the INTERSPEECH 2009 Emotion Challenge, organised by (part of) the authors, including the description of the Challenge's database, Sub-Challenges, participants and their approaches, the winners, and the fusion of results to the actual learnt lessons before we finally address the ever-lasting problems and future promising attempts.Keywords: emotion, affect, automatic classification, feature types, feature selection, noise robustness, adaptation, standardisation, usability, evaluation Setting the SceneThis special issue will address new approaches towards dealing with the processing of realistic emotions in speech, and this overview article will give an account of the state-of-the-art, of the lacunas in this field, and of promising approaches towards overcoming shortcomings in modelling and recognising realistic emotions. We will also report on the first emotion challenge at INTERSPEECH 2009, constituting the initial impetus of this special issue; to end with, we want to sketch future strategies and applications, trying to answer the question 'Where to go from here?'The article is structured as follows: we first deal with the basic phenomenon briefly reflecting the last fifteen years, commenting on databases, modelling and annotation, the unit of analysis and prototypicality. We then proceed to automatic processing (sec. 2) including discussions on features, classification, robustness, evaluation, and implementation and system integration. From there we go to the the first Emotion Challenge (sec. 3) including the description of the Challenge's database, Sub-Challenges, participants and their approaches, the winners, and the fusion of results to the lessons learnt, before concluding this article (sec. 4).

show abstract

“…Then, we can try to make better predictions using a larger ground truth or designing new audio descriptors especially relevant for this task. Another option would be to generate analytical features [17], to try to increase the accuracy of the system. Finally, the mood annotation could be personalized, learning from the user's feedback and his own perception of mood.…”

Section: Discussionmentioning

confidence: 99%

Music Mood Annotator Design and Integration

Laurier

Meyers

Serrà

et al. 2009

2009 Seventh International Workshop on Content-Based Multimedia Indexing

View full text Add to dashboard Cite

A robust and efficient technique for automatic music mood annotation is presented. A song's mood is expressed by a supervised machine learning approach based on musical features extracted from the raw audio signal. A ground truth, used for training, is created using both social network information systems and individual experts. Tests of 7 different classification configurations have been performed, showing that Support Vector Machines perform best for the task at hand. Moreover, we evaluate the algorithm robustness to different audio compression schemes. This fact, often neglected, is fundamental to build a system that is usable in real conditions. In addition, the integration of a fast and scalable version of this technique with the European Project PHAROS is discussed.

show abstract

Analytical Features: A Knowledge-Based Approach to Audio Feature Generation

Cited by 38 publications

References 61 publications

Indexing music by mood: design and integration of an automatic content-based annotator

Indexing music by mood: design and integration of an automatic content-based annotator

Recognising realistic emotions and affect in speech: State of the art and lessons learnt from the first challenge

Music Mood Annotator Design and Integration

Contact Info

Product

Resources

About