Exploring automatic music annotation with "acoustically-objective" tags

Abstract

The task of automatically annotating music with text tags (referred to as autotagging) is vital to creating a large-scale semantic music discovery engine. Yet for an autotagging system to be successful, a large and cleanly-annotated data set must exist to train the system. For this reason, we have collected a data set, called Swat10k, which consists of 10,870 songs annotated using a vocabulary of 475 acoustic tags and 153 genre tags from Pandora's Music Genome Project. The acoustic tags are considered "acoustically-objective" because they can be consistently applied to songs by expert musicologists. To develop an autotagging system, we use the Swat10k data set in conjunction with two new sets of content-based audio features obtained using the publiclyavailable Echo Nest API. The Echo Nest Timbre (ENT) features represent a song using a collection of short-time feature vectors. Compared with Mel-frequency cepstral coefficients (MFCCs), ENTs provide a more compact representation of music and improve autotagging performance. We also evaluate the Echo Nest Song (ENS) feature vector, which is a collection of mid-level acoustic features (e.g., beats per minute, average loudness). While the ENS features generally perform worse than the ENTs, they increase the performance of several individual tags. Furthermore, we plan to publicly release our song annotations and corresponding Echo Nest features so that other researchers will be able to use Swat10K to develop and compare alternative autotagging algorithms.

Singer identification using perceptual features and cepstral coefficients of an audio signal from Indian video songs

“…2 represents segmentation process from IVS. Echo Nest timbre, Echo Nest pitches (ENP), and Echo Nest loudness (ENL) are computed using Echo Nest API [15]. Segments are characterized by their perceptual onsets of an Indian song in Echo Nest API automatically.…”

“…Music researchers have started a company named the Echo Nest [15] in 2005. It gives free analysis of music via API.…”

A Method of Music Retrieval Using Impression Keywords

“…However, when the metadata cannot be used if the user doesn't know the title, this retrieval method cannot be applied. Therefore, there are studies related to the annotation [1][2] and the tagging [3], [4] in order to retrieve easily by enhancing the metadata. Moreover, there are researches which target music is retrieved by humming a part of the tune without using the metadata [5].…”

Virtual birding : extending birdwatching to review acoustic recordings

“…This is, in part, due to the time required to listen to recorded audio, contrasted to the time taken to interpret an image. In addition, automated techniques are not as advanced, and building a repository of annotations created by experts is not scalable and very expensive (Tingle, Kim, & Turnbull, 2010). There are relatively well-known attributes for images and text -for an image the dimensions, colour, composition style, and content; text is composed of groups of symbols with a recognisable font or style, which can be further grouped into words, sentences, and paragraphs.…”

“…Producing high-quality annotations can be done by employing birders dedicated to the task; however, this can be difficult to expand, due to financial restrictions or the time and effort required (Tingle et al, 2010).…”