Derek J. Shiell

High-power 280 nm AlGaN light-emitting diodes based on an asymmetric single-quantum well

We demonstrate high-power AlGaN-based ultraviolet light-emitting diodes grown on sapphire with an emission wavelength of 280 nm using an asymmetric single-quantum-well active layer configuration on top of a high-quality AlGaN/AlN template layer. An output power of 1.8 mW at a pulsed current of 400 mA was achieved for a single 300 μm×300 μm diode. This device reached a high peak external quantum efficiency of 0.24% at 40 mA. An array of four diodes produced 6.5 mW at 880 mA of pulsed current.

4.5 mW operation of AlGaN-based 267 nm deep-ultraviolet light-emitting diodes

We demonstrate 4.5 mW output power from AlGaN-based single quantum well ultraviolet light-emitting diodes at a very short wavelength of 267 nm in pulsed operation mode. The output power in continuous-wave mode reaches a value of 165 μW at an injected current of 435 mA. The measurements were done on arrays of four devices flip chip bonded to AlN submounts for thermal management.

High quantum efficiency AlGaN solar-blind p-i-n photodiodes

We report AlGaN-based back-illuminated solar-blind ultraviolet p-i-n photodetectors with a peak responsivity of 136 mA/W at 282 nm without bias. This corresponds to a high external quantum efficiency of 60%, which improves to a value as high as 72% under 5 V reverse bias. We attribute the high performance of these devices to the use of a very-high quality AlN and Al0.87Ga0.13N/AlN superlattice material and a highly conductive Si–In co-doped Al0.5Ga0.5N layer.

320×256 solar-blind focal plane arrays based on AlxGa1−xN

We report AlGaN-based backilluminated solar-blind ultraviolet focal plane arrays operating at a wavelength of 280 nm. The electrical characteristics of the individual pixels are discussed, and the uniformity of the array is presented. The p–i–n photodiode array was hybridized to a 320×256 read-out integrated circuit entirely within our university research lab, and a working 320×256 camera was demonstrated. Several example solar-blind images from the camera are also provided.

High-quantum-efficiency solar-blind photodetectors

We report AlGaN-based back-illuminated solar-blind p-i-n photodetectors with a record peak responsivity of 150 mA/W at 280 nm, corresponding to a high external quantum efficiency of 68%, increasing to 74% under 5 volts reverse bias. Through optimization of the p-AlGaN layer, we were able to remove the out-of-band negative photoresponse originating from the Schottky-like p-type metal contact, and hence significantly improve the degree of solar-blindness. We attribute the high efficiency of these devices to the use of very-high quality AlN and Al0.87Ga0.13N/AlN superlattice material, a highly conductive Si-In co-doped Al0.5Ga0.5N layer, and the elimination of the negative photoresponse through improvement of the p-type AlGaN.

Growth of deep-UV light-emitting diodes by metalorganic chemical vapor deposition

We demonstrate high power AlGaN based ultraviolet light-emitting diodes (UV LEDs) with an emission wavelength of 280 nm using an asymmetric single quantum well active layer configuration on top of a high-quality AlGaN/AlN template layer grown by metalorganic chemical vapor deposition (MOCVD). An output power of 1.8 mW at a pulsed current of 400 mA was achieved for a single 300 μm × 300 μm diode. This device reached a high peak external quantum efficiency of 0.24% at 40 mA. An array of four diodes produced 6.5 mW at 880 mA of pulsed current. We also demonstrate high output power operation of AlGaN-based UV LEDs at a short wavelength of 265 nm. An output power of 2.4 mW at a pulsed current of 360 mA was achieved for a single diode. A packaged array of four diodes produced 5.3 mW at 700 mA of pulsed current. The DC output power is 170 μW at 250 mA.

DNA Microarray Image Intensity Extraction using Eigenspots

DNA microarrays are commonly used in the rapid analysis of gene expression in organisms. Image analysis is used to measure the average intensity of circular image areas (spots), which correspond to the level of expression of the genes. A crucial aspect of image analysis is the estimation of the background noise. Currently, background subtraction algorithms are used to estimate the local background noise and subtract it from the signal. In this paper we use principal component analysis (PCA) to de-correlate the signal from the noise, by projecting each spot on the space of eigenvectors, which we term eigenspots. PCA is well suited for such application due to the structural nature of the images. To compare the proposed method with other background estimation methods we use the industry standard signal-to-noise metric xdev.

Vector quantization with memory and multi-labeling for isolated video-only automatic speech recognition

We describe a vector quantizer (VQ) with memory for automatic speech recognition (ASR) and compare the recognition performance results to those obtained with traditional memoryless VQ for ASR. Standard VQ for ASR quantizes the speech data independently of any past information. We introduce memory in a probabilistic framework for quantization state modeling. This is accomplished in the form of an ergodic hidden Markov model (HMM) in which the state occupied by the HMM represents the quantization label. We evaluate this approach in the context of video-only isolated digit ASR and implement both single stream (single labeling) and multi-stream (multi-labeling) systems. For single stream recognition, our approach increases the recognition rate from 62.67% to 66.95%. When using multi-labeling, our proposed vector quantizer with memory consistently outperforms the memoryless vector quantizer.

Sub-component modeling for face image reconstruction in video communications

Emerging communications trends point to streaming video as a new form of content delivery. These systems are implemented over wired systems, such as cable or ethernet, and wireless networks, cell phones, and portable game systems. These communications systems require sophisticated methods of compression and error-resilience encoding to enable communications across band-limited and noisy delivery channels. Additionally, the transmitted video data must be of high enough quality to ensure a satisfactory end-user experience. Traditionally, video compression makes use of temporal and spatial coherence to reduce the information required to represent an image. In many communications systems, the communications channel is characterized by a probabilistic model which describes the capacity or fidelity of the channel. The implication is that information is lost or distorted in the channel, and requires concealment on the receiving end. We demonstrate a generative model based transmission scheme to compress human face images in video, which has the advantages of a potentially higher compression ratio, while maintaining robustness to errors and data corruption. This is accomplished by training an offline face model and using the model to reconstruct face images on the receiving end. We propose a sub-component AAM modeling the appearance of sub-facial components individually, and show face reconstruction results under different types of video degradation using a weighted and non-weighted version of the sub-component AAM.