Ting, Wen YuanWen YuanTingWang, Syu SiangSyu SiangWangTsao, YuYuTsaoBORCHING SU2023-12-092023-12-092023-01-01979835032411221610363https://scholars.lib.ntu.edu.tw/handle/123456789/637637Beamforming techniques are popular in speech-related applications because of their effective spatial filtering capabilities. Nonetheless, conventional beamforming techniques generally depend on the target's direction-of-arrival (DOA), relative transfer function (RTF), or covariance matrix. This study presents a new approach, the intelligibility-aware null-steering (IANS) beamforming framework, which uses the STOI-Net intelligibility prediction model to improve speech intelligibility without prior knowledge of the aforementioned speech signal parameters. The IANS framework combines a null-steering beamformer (NSBF) to generate a set of beamformed outputs, and STOI-Net, to determine the optimal result. The experimental results indicate that the IANS framework can produce intelligibility-enhanced signals using a small dual-microphone array. The results are comparable to those obtained by null-steering beamformers with a given knowledge of the DOAs.beamforming | microphone arrays | null-steering | STOI | STOI-Netconference paper10.1109/MLSP55844.2023.102859132-s2.0-85177210329https://api.elsevier.com/content/abstract/scopus_id/85177210329