Audio encoding device and audio encoding method
An audio encoding device which can improve encoding performance while performing division search on an algebraic codebook in an audio encoding. In a distortion minimizing unit ( 112 ) of a CELP encoding device: a maximum correlation value calculation unit ( 221 ) calculates a correlation value by using each pulse and a target signal in each candidate position for four pulses constituting the fixed codebook so as to acquire a maximum value of the correlation value for each pulse and calculates a maximum correlation value by using the maximum value of the correlation value; a sorting unit ( 222 ) divides the four pulses into two subsets each having two pulses; and a search unit ( 224 ) performs a division search on the fixed codebook and acquires a code indicating the positions and polarities of the four pulses where the encoding distortion is minimum.
1. A speech encoding apparatus comprising:
a calculating section that calculates a correlation value for each candidate pulse position using a target signal and a plurality of pulses forming a fixed codebook, and calculates, on a per pulse basis, a representative scalar value of a pulse using a maximum value of the correlation values calculated for the pulse, wherein each representative scalar value corresponds to each pulse;
a sorting section that sorts the representative scalar values acquired on a per pulse basis, for the plurality of pulses, groups pulses corresponding to the sorted representative scalar values into a plurality of predetermined subsets and determines a first subset to be searched first and a second subset to be searched second among the plurality of subsets; and
a search section that searches the fixed codebook using the first subset and the second subset and acquires a code indicating positions and polarities of a plurality of pulses for minimizing coding distortion,
wherein the calculating section calculates, as the representative scalar value, the maximum correlation value of each pulse by adding a second highest correlation value multiplied by a predetermined rate to a maximum value of the correlation value on a per pulse basis.
2. The speech encoding apparatus according to claim 1 , wherein the sorting section sets a subset including a pulse corresponding to a highest representative scalar value among the representative scalar values acquired on a per pulse basis, as the first subset.
3. The speech encoding apparatus according to claim 1 , wherein:
the sorting section groups the pulses corresponding to the sorted representative scalar values into a plurality of combinations of a plurality of predetermined subsets, and determines the first subsets in the plurality of combinations, respectively; and the search section searches the fixed codebook using the first subsets and acquires the code to minimize the coding distortion.
4. The speech encoding apparatus according to claim 1 , wherein the sorting section determines the first subset using the representative scalar values corresponding to the grouped pulses.
5. The speech encoding apparatus according to claim 1 , wherein the sorting section generates a plurality of combinations of the representative scalar values corresponding to the grouped pulses, and determines the first subset based on a comparison result of the combinations multiplied by a predetermined value.
6. The speech encoding apparatus according to claim 1 , wherein the sorting section rearranges the pulses to be grouped into the plurality of subsets in a predetermined order.
7. A speech encoding method comprising the steps of:
calculating a correlation value for each candidate pulse position using a target signal and a plurality of pulses forming a fixed codebook, and calculating, on a per pulse basis, a representative scalar value of a pulse using a maximum value of the correlation values calculated for the pulse, wherein each representative scalar value corresponds to each pulse;
sorting the representative scalar values acquired on a per pulse basis, for the plurality of pulses, grouping pulses corresponding to the sorted representative scalar values into a plurality of predetermined subsets and determining a first subset to be searched first and a second subset to be searched second among the plurality of subsets; and
searching the fixed codebook using the first subset and the second subset and generating a code indicating positions and polarities of a plurality of pulses for minimizing coding distortion.