Sound Indexing Using Morphological Description

Abstract

Abstract-Sound sample indexing usually deals with the recognition of the source/cause that has produced the sound. For abstract sounds, sound-effects, unnatural or synthetic sounds this cause is usually unknown or unrecognizable. An efficient description of these sounds has been proposed by Schaeffer under the name morphological description. Part of this description consists in describing a sound by identifying the temporal evolution of its acoustic properties to a set of profiles. In this work, we consider three morphological descriptions: dynamic profiles (ascending, descending, ascending/descending, stable, impulsive), melodic profiles (up, down, stable, up/down, down/up) and complex-iterative sound description (non-iterative, iterative, grain, repetition). We study the automatic indexing of a sound into these profiles. Because this automatic indexing is difficult using standard audio features, we propose new audio features to perform this task. The dynamic profiles are estimated by modeling the loudness over-time of a sound by a second-order B-spline model and derive features from this model. The melodic profiles are estimated by tracking over time the perceptual filter which has the maximum excitation. A function is derived from this track which is then modeled using a second-order B-spline model. The features are again derived from the B-spline model. The description of complex-iterative sounds is obtained by estimating the amount of repetition and the period of the repetition. These are obtained by computing an audio similarity function derived from an MFCC similarity matrix. The proposed audio features are then tested for automatic classification. We consider three classification tasks corresponding to the three profiles. In each case, the results are compared with the ones obtained using standard audio features.

Improving Audio Retrieval through Loudness Profile Categorization

“…Though the source-centric description is important, as discussed earlier, a categorization in terms of perceptual traits (or morphological characteristics) would provide a more generic description for any kind of sound. In this context, we discuss Schaeffer's work [2] on typomorphology which has been utilized to build taxonomies for sound indexing and retrieval [3], [4].…”

“…Several descriptors have been explored in literature for quantification of these constructs and the given morphological components. Interested reader is referred to [6], [3], [4] for details on suggested taxonomies and representations. We use Peeters et al's [4] taxonomy with some modifications.…”

“…Interested reader is referred to [6], [3], [4] for details on suggested taxonomies and representations. We use Peeters et al's [4] taxonomy with some modifications. [7] We use [4], [3] as key references.…”

“…We use Peeters et al's [4] taxonomy with some modifications. [7] We use [4], [3] as key references. With respect to the previous work, our primary contributions are (i) proposal of a simple, intuitive and flexible approach for loudness profile categorization using thresholding (ii) demonstrating that filtering similarity search results based on query sound's loudness profile improves their relevance.…”

“…2. As indicated, for single events the loudness curve could belong to one of the following categories: impulsive if it has either a sharp attack or is of a very short duration; stable [4] if the loudness of the sound does not vary much; It is of increasing (or decreasing) category if the loudness increases (or decreases) for significant portion of the sound's duration. and of delta class if it is perceived of increasing-decreasing loudness.…”