Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | AE-P6.6 |
| Session: | Audio for Multimedia and Networks |
| Time: | Friday, May 21, 13:00 - 15:00 |
| Presentation: |
Poster |
| Topic: |
Audio and Electroacoustics: Network Audio |
| Title: |
FINE-GRANULAR SCALABLE AND ERROR RESILIENT AUDIO CODING BY TREE-STRUCTURED QUANTIZATION |
| Authors: |
Stian Johansen; Norwegian University of Science and Technology | | |
| | Andrew Perkis; Norwegian University of Science and Technology | | |
| | Tor Audun Ramstad; Norwegian University of Science and Technology | | |
| | Ajit Shyamsunder Bopardikar; Norwegian University of Science and Technology | | |
| Abstract: |
This paper presents a scheme for scalable, errorresilient coding of audio signals. The scheme is implemented hereas part of a subband-based audio codec and involves the union oftwo interconnected ideas: Tree-structured vector quantization(TSVQ) and reallocation of bits (RoB). The former ensures that the encoding procedure results in fixed-length codes while the latter is a low-complexity scaling algorithm that computes optimal bitrates for each subband when the bitstream needs to bedownscaled to a lower bitrate. The functionality enabled by thesetwo systems assumes great importance when the encoded material is to be delivered over channels and networks that have a time-varying bitrate, such as IP-based wireless networks. Suchchannels are often error-prone and the potentially debilitatingeffects of introduced errors are mitigated by the inherentfixed-length property of all codewords. We present results thathighlight the performance of this system under noisy conditionsand show that the degradation in quality is graceful. |
| |
| Back | |