Learn More
The convolutive blind source separation (BSS) problem can be solved efficiently in the frequency domain, where instantaneous BSS is performed separately in each frequency bin. However, the permutation ambiguity in each frequency bin should be resolved so that the separated frequency components from the same source are grouped together. To solve the(More)
  • Heping Ding
  • 2005
This paper proposes two patent-pending stable approaches to solve a symmetric positive-definite linear system associated with the fast affine projection (FAP) adaptation algorithm. The first approach can provide an exact solution at a relatively high complexity. The second, based on a novel reciprocating recursion algorithm, gives a reasonable approximate(More)
  • Heping Ding
  • 2004
This paper proposes a scheme of transmitting or storing a wideband audio stream in a channel or storage medium which is narrowband in nature and with a low resolution data format, such as an 8-bit companded one. In the proposed scheme, signal components in the upper-band part of the wideband beyond the narrowband - are encoded and then the code is embedded(More)
This paper proposes two noniterative approaches to solve a symmetric linear system associated with the fast affine projection adaptation algorithm. The first approach, using matrix LDL<sup>T</sup> factorization, can provide an exact solution at a moderate complexity, in contrast with the fact that existing stable approaches are all approximate. Based on a(More)
1. INTRODUCTION An echo canceller removes undesired echo in full-duplex speech communication. The cancellation is done by modeling the echo path impulse response with an adaptive finite impulse response filter and subtracting the echo estimate from the received signal. A typical diagram of an echo canceller is depicted in Figure 1. The signal x(n) and v(n)(More)
In general, it is difficult for conventional adaptive interference cancellation schemes to improve speech intelligibility in the presence of interference whose source is obtained asynchronously with the corrupted target speech. This is because there are inevitable timing drifts between the two inputs to the system. To address this problem, a(More)