초록 ▼

An apparatus for providing improved data compression may include an encoder comprising a quantizer for encoding input data and a side model. The quantizer may be trained with respect to high priority data among the input data and may be configured to partially encode the input data by encoding the h

An apparatus for providing improved data compression may include an encoder comprising a quantizer for encoding input data and a side model. The quantizer may be trained with respect to high priority data among the input data and may be configured to partially encode the input data by encoding the high priority data. The side model may be trained jointly with the training of the quantizer and is configured to model low priority data among the input data.

대표청구항 ▼

1. A method comprising: partially encoding, in a processor, input data including spectral envelope data and residual spectrum data by encoding high priority data among the input data using a quantizer trained with respect to the high priority data, wherein the high priority data is associated with t

1. A method comprising: partially encoding, in a processor, input data including spectral envelope data and residual spectrum data by encoding high priority data among the input data using a quantizer trained with respect to the high priority data, wherein the high priority data is associated with the spectral envelope data of the input data; andmodeling, in a processor, low priority data among the input data using a side model trained jointly with the training of the quantizer, wherein the low priority data is associated with the residual spectrum data of the input data, andwherein partially encoding the input data comprises providing a selector matrix to enable encoding only the input data that corresponds to the high priority data. 2. The method of claim 1, further comprising decoding compressed data using a dequantizer and the side model to recover the high priority data and the low priority data, respectively, from a received transmission. 3. The method of claim 1, wherein partially encoding the input data comprises compressing the input data without using any bits for quantization of the low priority data. 4. The method of claim 1, further comprising providing an indicator bit indicative of a sufficiency of the accuracy of the modeling of the low priority data and providing modeling error information for improving modeling accuracy in response to the indicator bit indicating insufficient modeling accuracy. 5. The method of claim 1, wherein partially encoding the input data comprises quantizing the high priority data using the quantizer that is further trained to enhance quantization accuracy for the high priority data while simultaneously training the side model; andthe low priority data is not quantized and not encoded. 6. A computer program product comprising at least one computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising: a first executable portion for partially encoding input data including spectral envelope data and residual spectrum data by encoding high priority data among the input data using a quantizer trained with respect to the high priority data, wherein the high priority data is associated with the spectral envelope data of the input data; anda second executable portion for modeling low priority data among the input data using a side model trained jointly with the training of the quantizer, wherein the low priority data is associated with the residual spectrum data of the input data, andwherein the first executable portion includes instructions for providing a selector matrix to enable encoding only the input data that corresponds to the high priority data. 7. The computer program product of claim 6, further comprising a third executable portion for decoding compressed data using a dequantizer and the side model to recover the high priority data and the low priority data, respectively, from a received transmission. 8. The computer program product of claim 6, wherein the first executable portion includes instructions for compressing the input data without using any bits for quantization of the low priority data. 9. The computer program product of claim 6, further comprising a third executable portion for providing an indicator bit indicative of a sufficiency of the accuracy of the modeling of the low priority data and a fourth executable portion for providing modeling error information for improving modeling accuracy in response to the indicator bit indicating insufficient modeling accuracy. 10. The computer program product of claim 6, wherein the first executable portion includes instructions for quantizing the high priority data using the quantizer that is further trained to enhance quantization accuracy for the high priority data while simultaneously training the side model; and,the low priority data is not quantized and not encoded. 11. An apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to: partially encode image input data including spectral envelope data and residual spectrum data by encoding high priority data among the input data using a quantizer trained with respect to the high priority data, wherein the high priority data is associated with the foreground data of the input signal, andmodel low priority data among the input data using a side model trained jointly with the training of the quantizer, wherein the low priority data, is associated with the residual spectrum data of the image input data, andwherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the apparatus at least to provide a selector matrix to enable encoding only the input data that corresponds to the high priority data. 12. The apparatus of claim 11, wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the apparatus at least to decode compressed data using a dequantizer and the side model to recover the high priority data and the low priority data, respectively, from a received transmission. 13. The apparatus of claim 11, wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the apparatus at least to compress the input audio data without using any bits for quantization of the low priority data. 14. The apparatus of claim 11, wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the apparatus at least to provide an indicator bit indicative of a sufficiency of the accuracy of the modeling of the low priority data and to cause the apparatus at least to receive modeling error information for improving modeling accuracy in response to the indicator bit indicating insufficient modeling accuracy. 15. The apparatus of claim 11, wherein the at least one memory and the computer program code are further configured to, with the at least one processor, cause the apparatus at least to quantize the high priority data using the quantizer that is further trained to enhance quantization accuracy for the high priority data while simultaneously training the side model; and,the low priority data is not quantized and not encoded.