Entity

Time filter

Source Type

Kwai, Hong Kong

Yuan J.,Hong Kong University of Science and Technology | Fung S.W.,Technology Research Institute Co. | Chan K.Y.,ASM Pacific Technology Ltd. | Xu R.,Hong Kong University of Science and Technology
IEEE Transactions on Instrumentation and Measurement | Year: 2012

The linearity of a pipeline analog-to-digital converter (ADC) is mainly limited by capacitor mismatch and finite operational amplifier (OPAMP) gain, which cause large power and design difficulty in modern nanometer CMOS processes for high-resolution pipeline ADCs. It is a trend of developing digital calibration techniques to compensate the analog error in pipeline stages. This paper systematically introduces a novel interpolation-based digital calibration architecture to compensate both linear and nonlinear errors from pipeline stages. The new method does not require convergence. The effect of calibration error is analyzed in detail in this paper. A prototype 20-MS/s pipeline ADC is fabricated in a 0.35-μm 3.3-V CMOS process. For 12-b resolution, the digital calibration improves the ADC differential nonlinearity and integral nonlinearity from 1.47 LSB and 7.85 LSB to 0.2 LSB and 0.27 LSB. For a 590-kHz sinusoidal signal, the calibration improves the ADC signal-to-noise-distortion ratio and spurious-free dynamic range from 41.3 dB and 52.1 dB to 72.5 dB and 84.4 dB, respectively. With the new calibration technique, low-gain OPAMPs and small capacitors are used in the pipeline. The designed ADC has 0.78-pJ/step figure of merit (FOM), which is among the lowest reported FOMs for high-resolution pipeline ADC designs. The new architecture requires an accurate calibration ADC (CalADC) and two digital decoders. CalADC is implemented on-chip with 6.5% die area and 8.9% power. The decoders are synthesized to have 912 gates and consume 23.4% ADC power. © 2011 IEEE. Source


Chan F.Y.M.,ASM Pacific Technology Ltd. | Mudhana G.,Vellore Institute of Technology
Journal of the Optical Society of America A: Optics and Image Science, and Vision | Year: 2014

We investigate how the index profile of a few-mode fiber (FMF) can be designed so that group velocities of the two lowest-order modes can be equalized at a normalized frequency, which is below the cut-off frequency of the LP 21 mode. This can be achieved using a single-clad power-law profile with a sufficiently large profile exponent or a double-clad profile consisting of a graded-core surrounded by a sufficiently thick depressed inner cladding without index jump at their interface. The fabrication tolerances, effective index differences, intramodal dispersion differences, and effective mode areas of various single- and double-clad profiles are compared. The results show that, in comparison to single-clad fibers, double-clad fibers are capable of producing higher fabrication tolerances and reduced sensitivity of group delay difference to wavelength by three and two orders of magnitude, respectively. Our analyses provide insights into the design of FMFs, which will facilitate future development of highcapacity mode division long-haul transmission systems. © 2014 Optical Society of America. Source


Bai T.,ASM Pacific Technology Ltd. | Li Y.,City University of Hong Kong | Zhou X.,Zhejiang University of Technology
Advanced Robotics | Year: 2014

In this work, we address the problem of monocular tracking the human motion based on the discriminative sparse representation. The proposed method jointly trains the dictionary and the discriminative linear classifier to separate the human being from the background. We show that using the online dictionary learning, the tracking algorithm can adapt the variation of human appearance and background environment. We compared the proposed method with four state-of-the-art tracking algorithms on eight benchmark video clips (Faceocc, Sylv, David, Singer, Girl, Ballet, OneLeaveShopReenter2cor, and ThreePastShop2cor). Qualitative and quantitative experimental validation results are discussed at length. The proposed algorithm for human tracking achieves superior tracking results, and a Matlab run time on a standard desktop machine of four frames per second. © 2014 Taylor & Francis and The Robotics Society of Japan. Source


Qin J.,Smart China Holdings Ltd | Deng F.,ASM Pacific Technology Ltd. | Yung N.H.C.,University of Hong Kong
Signal, Image and Video Processing | Year: 2014

With the bag-of-contextual-visual-word (BOCVW) models, we propose a scene categorization method based on local–global feature fusion and multi-scale multi-spatial resolution encoding. First, the performances of the BOCVW models belonging to different features are mutually reinforced by fusing other types of features within local regions. Then, the spatial configuration information is explored using a multi-scale multi-spatial resolution encoding approach. Furthermore, these encoded BOCVW models are globally fused using an improved maximum-margin optimization strategy, which considers the margin between input vectors of different categories and the diameter of the smallest ball containing feature vectors simultaneously. The proposed method has been evaluated on three scene categorization datasets consisting of scene categories 8, 15, and 67, respectively. And its effectiveness has been verified by these experimental results. © 2014, Springer-Verlag London. Source


Chan F.Y.M.,ASM Pacific Technology Ltd. | Mudhana G.,Vellore Institute of Technology | Shum P.,Nanyang Technological University
Applied Optics | Year: 2015

We present a comparative study of the bandwidth and the sensitivity of the resonance wavelength of long-period gratings (LPGs) to external perturbation fabricated in single-mode fibers (SMFs) and few-mode fibers (FMFs), and their dependencies on the group indices and the dispersion properties of the phase-matched modes. Unlike SMFs, a relatively large core size of FMFs invariably leads to nonuniform index modulation across the fiber cross section under UV exposure, enabling the coupling between modes having dissimilar azimuthal symmetry. Simple analytical formulas for the group/effective index difference, dispersion difference, bandwidth, and wavelength sensitivities are derived for the case of SMFs where light is coupled from the fundamental core mode to the symmetrical cladding modes. Our results show that a two-mode fiber operating at a V -number close to 3 is capable of producing LPGs with broader bandwidth and higher sensitivity as compared with their SMF counterparts, except for a few special cases. Our analyses provide insights into the characteristics of LPGs and facilitate their designs for specific applications. © 2015 Optical Society of America. Source

Discover hidden collaborations