| 摘要 |
<p>A data processing apparatus capable of improving the quality of decoded data on an image or speech. A decoder (21), for example, decodes encoded data created by JPEG-encoding an image. An encoder (22) encodes first decoded data created by the decoding and outputs re-encoded data. Furthermore, a decoder (23) decodes the re-encoded data and produces second decoded data. A classification adaptive processing circuit (24) generates from the first decoded data and the second decoded data a prediction tap for performing a predetermined prediction calculation with a tap coefficient determined by learning by using the prediction tap and the tap coefficient, thereby determining a prediction value corresponding to teacher data which has been used as a teacher in the learning. The present invention can be applied to a transmission system transmitting an encoded image and voice.</p> |
