Audio thumbnailing of popular music using chroma-based representations

Abstract

Abstract-With the growing prevalence of large databases of multimedia content, methods for facilitating rapid browsing of such databases or the results of a database search are becoming increasingly important. However, these methods are necessarily media dependent. We present a system for producing short, representative samples (or "audio thumbnails") of selections of popular music. The system searches for structural redundancy within a given song with the aim of identifying something like a chorus or refrain. To isolate a useful class of features for performing such structure-based pattern recognition, we present a development of the chromagram, a variation on traditional time-frequency distributions that seeks to represent the cyclic attribute of pitch perception, known as chroma. The pattern recognition system itself employs a quantized chromagram that represents the spectral energy at each of the 12 pitch classes. We evaluate the system on a database of popular music and score its performance against a set of "ideal" thumbnail locations. Overall performance is found to be quite good, with the majority of errors resulting from songs that do not meet our structural assumptions.

Voice activity detection algorithm based on long-term pitch information

“…The decomposition is realized by a suitable multirate filter bank consisting of elliptic filters [13]. This representation of audio signals can then be used as a basis for deriving various audio features of various characteristics [13,14], such as chroma pitch, chroma log pitch, and chroma energy normalized statics [15]. Figure 1 shows the waveform of an audio recording and its corresponding pitch features in various noise environments.…”

Unsupervised Discovery of Temporal Structure in Music

“…Another application is that of music thumbnailing [17], which seeks to automatically identify a representative excerpt from a song. Utilizing the analysis described in this section, the thumbnail that best represents the identified motif (typically a good thumbnail since it is the most often repeated pattern found in the song) simply corresponds to the largest activation in H. If a longer thumbnail is required, then L or the parameters of α wτ can be scaled up to identify longer patterns.…”

“…This property has been exploited for tasks as diverse as visualization, rhythmic analysis, automatic summarization and thumbnailing, chorus detection, annotation, synchronization and long-term segmentation [16]- [20]. With a few exceptions [17], [21], [22], the emphasis of this research has been on locating repetitions rather than on extracting of characteristic, repetitive patterns. The utility of extracting such patterns is illustrated by previous research on detecting motif occurrences across a collection [23] and cover-song retrieval based on feature sub-sequences [24].…”

Sparse chroma estimation for harmonic non-stationary audio

“…Chroma features are widely used in applications such as cover song detection, transcription, and recommender systems (see, e.g. [7][8][9]). Most methods for chroma estimation begin with some pitch estimation, which then maps into its respective chroma.…”