| 摘要 |
The application relates to sound source separation in a hearing device. The hearing device comprises a cyclic analysis buffer unit of length A adapted for storing the last A audio samples, a cyclic synthesis buffer unit of length L (L<A), and a database having stored recorded sound examples ("atoms") from two or more sound sources (S 1 , S 2 ). The database comprises an analysis dictionary (A 1 and A 2 ) containing atoms of length A, and a reconstruction dictionary (R 1 and R 2 ) containing atoms of length L. Each atom in the analyis dictionary corresponds to a particular atom in the reconstruction dictionary. A sound source separation (SSU) unit is configured to determine the most optimal representation (W) of the last A samples given the atoms in the analysis dictionary, and to generate said at least two sound sources by combining atoms in the reconstruction dictionary using the optimal representation (W). The invention may e.g. be used for hearing devices, e.g. hearing aids, headsets, ear phones, active ear protection systems, handsfree telephone systems, mobile telephones, teleconferencing systems, public address systems, classroom amplification systems, etc. |
