Speech Enhancement by Adaptive Noise Cancellation: Problems, Algorithms, and Limits
"In this paper, a summary about adaptive noise reduction techniques for enhancing degraded audio signals in the context of forensic application is given. The surrounding conditions for the use of adaptive systems are examined at first to understand the functionality of the algorithms that are commonly employed. Built on this analyses we will give examples of self-adaptive single-channel solutions, monophonic and stereo noise-cancellers with reference signal and solutions for GSM burst removal. Furthermore, we investigate the weaknesses of the standard approaches and highlight problems that are resulting. Several difficulties generally have to be taken care of in the field of adaptive noise cancellation approaches to those are presented, but also the limits of feasibility."
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is temporarily free for AES members.