We present the theoretical modeling and prototype demonstration of a three-dimensional broadband, low-loss, dual-polarization, and high-directivity lens antenna using gradient index (GRIN) metamaterials, which is composed of multi-layer microstrip square-ring arrays. The elements of metamaterials, closed square-ring units of variable sizes, are distributed on the planar substrate to satisfy the radial gradient index function and the axial impedance matching layer configuration of the lens. The gradient-index metamaterials are designed to transform the spherical wave-front into the planar wave-front and to minimize the reflection loss. A prototype lens antenna, which consists of a metal conical horn and the gradient-index lens, are simulated, constructed, and measured. The resemblance of simulation and measurement results shows that the prototype lens antenna maintains low return loss and high directivity on the whole X-band (from 8 GHz to 12 GHz). Compared to the traditional horn antenna, the metamaterial GRIN lens antenna has much superior performance—for instance, the gain increases by 6 dBi at 12 GHz. These results demonstrate the feasibility of such a light weight slab metamaterial lens for broadband and high-directivity antenna applications, such as in radar and communication systems. We have used the lens antennas in the measurements of a three-dimensional invisibility cloak due to the high directivity.

No abstract

No abstract

Even though speaker recognition has gained significant progress in recent years, its performance is known to be deteriorated severely with the existence of strong background noises. Inspired by a recently proposed clean-frame selection approach, this work investigates a relatively elegant weighting method when computing the Baum-Welch statistics of Gaussian mixture models (GMMs) in i-vector extraction. By introducing weighting parameters to the frames of enrollment/testing utterances, the optimization problem is redefined and solved. New updating rules are derived by incorporating weights to the computation of posterior probabilities, mean vectors, and covariance matrices of the GMM. The experiments conducted on the Speakers in the Wild (SITW) database show that the proposed algorithm has significantly improved the performance of i-vector-based speaker recognition systems in noisy environments. Compared with the GMM i-vector baseline, the equal error rate is reduced from 5.75 to 4.72 and the minimum value of cost function (C min det) is reduced from 0.4825 to 0.4505. Slight but significant superiority is also observed over the method with an additional feature enhancement frontend by using deep neural networks. INDEX TERMS Gaussian mixture models, frame weighting, Baum-Welch statistics, i-vector, robust speaker recognition.

Even though audio replay detection has improved in recent years, its performance is known to severely deteriorate with the existence of strong background noises. Given the fact that different frames of an utterance have different impacts on the performance of spoofing detection, this paper introduces attention-based long short-term memory (LSTM) to extract representative frames for spoofing detection in noisy environments. With this attention mechanism, the specific and representative frame-level features will be automatically selected by adjusting their weights in the framework of attention-based LSTM. The experiments, conducted using the ASVspoof 2017 dataset version 2.0, show that the equal error rate (EER) of the proposed approach was about 13% lower than the constant Q cepstral coefficients-Gaussian mixture model (CQCC-GMM) baseline in noisy environments with four different signal-to-noise ratios (SNR). Meanwhile, the proposed algorithm also improved the performance of traditional LSTM on audio replay detection systems in noisy environments. Experiments using bagging with different frame lengths were also conducted to further improve the proposed approach.