Combining audio content and social context for semantic music discovery

Turnbull, Douglas; Barrington, Luke; Lanckriet, Gert R. G.; Yazdani, Mehrdad

doi:10.1145/1571941.1572009

Cited by 45 publications

(31 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also want to compare the performance of the proposed method with that of the conventional methods [9,11] and apply to various challenging real-world problems e.g., multi-modal event correlation anal- …”

Section: Discussionmentioning

confidence: 99%

“…To cope with this problem, most previous works have tried to automatically associate sounds with words for queryby-text retrieval or music annotation [4,5,6,7,8,9,10,11]. Recently, inference techniques based on topic models, such as probabilistic latent semantic analysis (pLSA) and latent Dirichlet allocation (LDA), have been exploited for automatic image annotation and retrieval [12,13].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic audio tag classification via semi-supervised canonical density estimation

Takagi

Ohishi

Kimura

et al. 2011

2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We propose a novel semi-supervised method for building a statistical model that represents the relationship between sounds and text labels ("tags"). The proposed method, named semi-supervised canonical density estimation, makes use of unlabeled sound data in two ways: 1) a low-dimensional latent space representing topics of sounds is extracted by a semi-supervised variant of canonical correlation analysis, and 2) topic models are learned by multi-class extension of semi-supervised kernel density estimation in the topic space. Real-world audio tagging experiments indicate that our proposed method improves the accuracy even when only a small number of labeled sounds are available.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Automatic audio tag classification via semi-supervised canonical density estimation

Takagi

Ohishi

Kimura

et al. 2011

2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…"happy" or "rock") to refer to music. Semantic/tag-based or category-based retrieval systems such as the ones proposed by Knees et al [125] or Turnbull et al [278] rely on methods for the estimation of semantic labels from music. This retrieval scenario is characterized by a low specificity and long-term granularity.…”

Section: Music Retrievalmentioning

confidence: 99%

Music Information Retrieval: Recent Developments and Applications

Schedl

Gómez

Urbano

2014

FNT in Information Retrieval

171

108

View full text Add to dashboard Cite

“…Few methods have been proposed that combine the information from both acoustic content and social context. In [13] timbre and harmonic features are used to represent acoustic content while social tags and web documents represent social context. Similar combining approaches are used in [14] for multi-label music style classification and in [15] where a track's significant musical content or musword is considered along with social tags.…”

Section: Related Workmentioning

confidence: 99%

Multi-label annotation of music

Ahsan

Kumar

Jawahar

2015

2015 Eighth International Conference on Advances in Pattern Recognition (ICAPR)

View full text Add to dashboard Cite

Abstract-Automatic annotation of an audio or a music piece with multiple labels helps in understanding the composition of a music. Such meta-level information can be very useful in applications such as music transcription, retrieval, organization and personalization. In this work, we formulate the problem of annotation as multi-label classification which is considerably different from that of a popular single (binary or multi-class) label classification. We employ both the nearest neighbour and max-margin (SVM) formulations for the automatic annotation. We consider K-NN and SVM that are adapted for multi-label classification using one-vs-rest strategy and a direct multi-label classification formulation using ML-KNN and M3L. In the case of music, often the signatures of the labels (e.g. instruments and vocal signatures) are fused in the features. We therefore propose a simple feature augmentation technique based on non-negative matrix factorization (NMF) with an intuition to decompose a music piece into its constituent components. We conducted our experiments on two data sets -Indian classical instruments dataset and Emotions dataset [1], and validate the methods.

show abstract

Combining audio content and social context for semantic music discovery

Cited by 45 publications

References 20 publications

Automatic audio tag classification via semi-supervised canonical density estimation

Automatic audio tag classification via semi-supervised canonical density estimation

Music Information Retrieval: Recent Developments and Applications

Multi-label annotation of music

Contact Info

Product

Resources

About