Kishore Kotteri

Low bitrate audio coding using generalized adaptive gain shape vector quantization across channels

Audio coding at low bitrates suffers from artifacts due to spectrum truncation. Typical audio codecs code multi-channel sources using transforms across the channels to remove redundancy such as middle (mid) - side (M/S) coding. At low bitrates, the spectrum of the coded channels is truncated and the spectrum of the channels with lower energy, such as the side channel, is truncated severely, sometimes entirely. This results in a muffled sound due to truncation of all coded channels beyond a certain frequency. It also results in a loss of spatial image even at low frequencies due to severe truncation of the side channel. Previously we have developed a low bitrate coding method to combat the loss of higher frequencies caused by spectrum truncation. In this paper, we present a novel low bitrate audio coding scheme to mitigate the loss of spatial image. Listening tests show that the combination of the two low bitrate coding methods results in a audio codec that can get good quality even at bitrates as low as 32kbps for stereo content with low decoder complexity.

Multiplierless filter Bank design: structures that improve both hardware and image compression performance

Design techniques for high-performance, fixed-point, multiplierless filter banks are presented. Image compression using the biorthogonal 9/7 discrete wavelet transform provides a motivating example. Image compression and hardware performance of two commonly used filter structures, direct and cascade, and two known filter bank structures, nonpolyphase and polyphase, are compared. A technique is shown for designing a fixed-point polyphase filter structure, which is highly efficient from a hardware standpoint, such that image-compression quality is not significantly deteriorated by the use of fixed-point mathematics. The result is a polyphase structure with about twice the throughput rate of nonpolyphase structures, and peak signal-to-noise ratio values for lossy compression within 0.2 decibels of those achieved using floating-point filters.