| 摘要 |
<p>The present invention relates to an intelligibility-enhancing apparatus and a voice output apparatus using same. An input envelope detection unit detects the level of a voice frame for an input signal, and provides the detected level to a cutoff frequency estimation unit. An output envelope detection unit detects the level of a voice frame for an output signal, and provides the detected level to a cutoff frequency estimation unit. The cutoff frequency estimation unit detects a difference value between the level of the Nth voice frame received from the input envelope detection unit and the level of the (N-1)th voice frame received from the output envelope detection unit, and calculates a cutoff frequency for amplifying the consonant component of the input signal using the difference value. A shelving filter unit filters the input signal in accordance with the cutoff frequency calculated by the cutoff frequency estimation unit, such that a portion estimated as the consonant component of the input signal is selectively amplified. Consequently, the shelving filter unit enables the cutoff frequency to dynamically vary in accordance with the quantity of consonant components of the input signal, and thus amplifies a high frequency component corresponding to the resonant component of the input signal while attenuating a low frequency component. Thus, a consonant component is selectively emphasized to enhance speech intelligibility.</p> |
