Semantic audio analysis has become a fundamental task in contemporary audio applications; consequently, further improvement and optimization of classification algorithms has also become a necessity. During the recent years, standard frame-based audio classification methods have been optimized and modern approaches introduced additional feature engineering steps, attempting to capture the temporal dependency between successive feature observations. This type of processing is known as Temporal Feature Integration. In this paper, the enhancement of statistical feature integration is proposed by extending and extensively evaluating the measures that can be deployed. Under this scope, new functions for capturing the shape of a texture window are introduced and evaluated. The ultimate goal of this work is to highlight the best performing measures for early temporal integration, focusing on simple feature engineering, avoiding complexity, and forming a compact and robust set of meta-features that can improve performance in audio classification tasks.
Click to purchase paper as a non-member or you can login as an AES member to see more options.
No AES members have commented on this paper yet.
To be notified of new comments on this paper you can
subscribe to this RSS feed.
Forum users should login to see additional options.
If you are not yet an AES member and have something important to say about this paper then we urge you to join the AES today and make your voice heard. You can join online today by clicking here.
