Towards Music Captioning: Generating Music Playlist Descriptions

Proc. ICMC, 2002

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

One of the main limitations in the field of audio signal processing is the lack of large public datasets with audio representations and high-quality annotations due to restrictions of copyrighted commercial music. We present Melon Playlist Dataset, a public dataset of mel-spectrograms for 649,091 tracks and 148,826 associated playlists annotated by 30,652 different tags. All the data is gathered from Melon, a popular Korean streaming service. The dataset is suitable for music information retrieval tasks, in particular, auto-tagging and automatic playlist continuation. Even though the latter can be addressed by collaborative filtering approaches, audio provides opportunities for research on track suggestions and building systems resistant to the cold-start problem, for which we provide a baseline. Moreover, the playlists and the annotations included in the Melon Playlist Dataset make it suitable for metric learning and representation learning.

2007

Large music collections afford the listener flexibility in the form of choice, which enables the listener to choose the appropriate piece of music to enhance or complement their listening scenario on-demand. However, bundled with such a large music collection is the daunting task of manually searching through each entry in the collection to find the appropriate song required by the listener. This often leaves the listener frustrated when trying to select songs from a large music collection. In this paper, an overview of existing methods for automatically generating a playlist is discussed. This discussion outlines advantages and disadvantages associated with such implementations.

Proceedings of 25th International AES Conference London, 2004

With the rapid growth of audio databases, many music retrieval applications have employed metadata descriptions to facilitate better handling of huge databases. Music structure creates the uniqueness identity for each music piece. Therefore, structural description is capable of providing a powerful way of interacting with audio content, and serves as a linkage between low-level description and higher-level descriptions of audio (eg audio summarization, audio fingerprinting, etc.). Identification of representative musical excerpts ...

2011

Abstract Music systems that generate playlists are gaining increasing popularity, yet ways to select songs to be acceptable to users is still elusive. We present the results of an explorative study that focused on the language of musically untrained end users for playlist choices, in a variety of listening contexts. Our results indicate that there are a number of opportunities for playlist recommendation or retrieval systems, particularly by taking context into account.

2005

We report experiments on the use of standard natural language processing (NLP) tools for the analysis of music lyrics. A significant amount of music audio has lyrics. Lyrics encode an important part of the semantics of a song, therefore their analysis complements that of acoustic and cultural metadata and is fundamental for the development of complete music information retrieval systems. Moreover, a textual analysis of a song can generate ground truth data that can be used to validate results from purely acoustic methods. Preliminary results on language identification, structure extraction, categorization and similarity searches suggests that a lot of profit can be gained from the analysis of lyrics.

2011

The Worldwide Web/Internet has changed the music industry by making huge amount of music available to both music publishers and consumers including ordinary listeners or end users. The Web2.0 tagging techniques of music items by artist name, album title, musical style or genre (technically these are termed as syntactic metadata) have given rise to the generation unstructured free form vocabularies. Music search based on these syntactic metadata requires the search query to contain at least one keyword from that vocabulary and it must be an exact match. The semantic Web initiative by W3C proposes machine process-able representation of information but does not stipulate how that can be applied to music items specifically. In this paper we present a novel approach that details a semi-automatic semantic annotation tool to enable music producers to generate music metadata through a mapping between music consumers' free form tags and the acoustic metadata that are automatically extractable from music audio. The proposed annotation tool enables onotology guided annotation process and uses MPEG-7 Audio compliant music annotation ontology represented in dominant semantic web standard OWL 1.0.

2019

Techniques of this disclosure may enable a computing device to suggest one or more titles based on the content of audio being recorded or audio that was previously recorded, and other data such as time and location. Rather than applying a general default title or audio file name, the computing device may request authorization from a user to analyze the contents of a recorded audio file and, after receiving explicit authorization from the user, analyze the audio, including speech, and automatically suggest titles that are indicative of the content of the audio and/or other data. The computing device may convert speech included in the audio into text and extract a plurality of terms from the text based on various factors, such as word classes (e.g., convert audio that includes “this meatball recipe adds parmesan cheese” into text and extract a plurality of nouns such as “meatball,” “recipe,” “parmesan,” and “cheese” from the text). Based on various factors, such as term frequency in t...

International Journal of Database Theory and Application, 2016

In this research work automatic tagging of songs using machine learning has been performed so that searching could be made effective while selecting songs. The goal of this research paper is to propose a system that will automatically recognize the genre of the tracks and tag them respectively by using different parameters obtained by acoustic analysis. The research work utilizes different combinations of algorithms and music parameters to accurately classify tracks into genres.

Journal of Multimedia, 2009

IEEE 1599 is a new XML-based format to describe heterogeneous music contents comprehensively. In a single file, music symbols, printed scores, audio tracks, computer-driven performances, catalogue metadata, text and graphic contents related to a single music piece are linked and mutually synchronized within the same framework. Heterogeneous contents are organized in a multilayered structure that supports different encoding formats and