Acoustic reflector localization is an important issue in audio signalprocessing, with direct applications in spatial audio, scene reconstruction,and source separation. Several methods have recently been proposed to estimatethe 3D positions of acoustic reflectors given room impulse responses (RIRs). Inthis article, we categorize these methods as "image-source reversion", whichlocalizes the image source before finding the reflector position, and "directlocalization", which localizes the reflector without intermediate steps. Wepresent five new contributions. First, an onset detector, called the clustereddynamic programming projected phase-slope algorithm, is proposed toautomatically extract the time of arrival for early reflections within the RIRsof a compact microphone array. Second, we propose an image-source reversionmethod that uses the RIRs from a single loudspeaker. It is constructed bycombining an image source locator (the image source direction and range (ISDAR)algorithm), and a reflector locator (using the loudspeaker-image bisection(LIB) algorithm). Third, two variants of it, exploiting multiple loudspeakers,are proposed. Fourth, we present a direct localization method, the ellipsoidtangent sample consensus (ETSAC), exploiting ellipsoid properties to localizethe reflector. Finally, systematic experiments on simulated and measured RIRsare presented, comparing the proposed methods with the state-of-the-art. ETSACgenerates errors lower than the alternative methods compared through ourdatasets. Nevertheless, the ISDAR-LIB combination performs well and has a runtime 200 times faster than ETSAC.
展开▼
