A novel approach for remotely sensing mechanical cardiovascular activity for use as a biometric marker is proposed. Laser Doppler Vibrometry (LDV) is employed to sense vibrations on the surface of the skin above the carotid artery related to arterial wall movements associated with the central blood pressure pulse. Carotid LDV signals are recorded using noncontact methods and the resulting unobtrusiveness is a major benefit of this technique. Several recognition methods are proposed that use the temporal and/or spectral information in the signal to assess biometric performance both on an intrasession basis, and on an intersession basis where LDV measurements were acquired from the same subjects after delays ranging from one week to six months. A waveform decomposition method that utilizes principal component analysis is used to model the signal in the time domain. Authentication testing for this approach produces an equal-error rate of 0.5% for intrasession testing. However, performance degrades substantially for intersession testing, requiring a more robust approach to modeling. Improved performance is obtained using techniques based on time-frequency decomposition, incorporating a method for extracting informative components. Biometric fusion methods including data fusion and information fusion are applied to train models using data from multiple sessions. As currently implemented, this approach yields an intersession equal-error rate of 6.3%.

We study the influence of the variation in the shape of the pits for the two-dimensional optical storage (TwoDOS) paradigm vis-à-vis the readback signal and decoding performance. The algorithm used for decoding is a message-passing based joint equalization and decoding algorithm proposed by Singla and O'Sullivan (Intl. Symp. Info. Theory, Sept. 2005). The performance dependence of the algorithm on pit shape is studied by comparing the noise-tolerance thresholds for the algorithm. The TwoDOS model under consideration also includes two forms of media noise: pit-position noise, arising due to the deviation in the location of the pits from their intended positions and pit-size noise, arising due to variations in the radii of the pits. It is shown that as the shape of the pit changes from the ideal cylindrical shape towards a conical shape there is a sharp decrease in the range of the signal intensity. This decrease in the range leads to increased signal folding which is detrimental for the decoding algorithm. However, we show that when the ratio of the area of the pit to the area of the hexagonal bit cell is large (~1) then the extent of signal folding for a conical profile is actually less than that of a cylindrical profile. Then, in this case, the performance of the decoder for a conical profile is better than for a cylindrical profile.

In this paper we examine the achievable rate region for the problem of successive refinement of information for pattern recognition systems. The pattern recognition system has two stages, going from coarse to fine recognition as more resources become available for storing internal representations of the patterns. We present an inner and an outer bound on the true achievable rate region. Using these results we derive conditions under which a pattern recognition system is successively refinable. These conditions are similar to the Markov condition for successive refinement in the rate-distortion problem.

An algorithm that performs joint equalization and decoding for channels with nonlinear two-dimensional intersymbol interference is presented. The algorithm performs sum-product message-passing on a factor graph that represents the underlying system. The two-dimensional optical storage (Two-DOS) technology is an example of a system with nonlinear two-dimensional intersymbol interference. Simulations for the nonlinear channel model of TwoDOS show significant improvement in performance over uncoded performance. Noise tolerance thresholds for the TwoDOS channel computed using density evolution are also presented

Joint iterative equalization and decoding schemes are proposed for two-dimensional inter-symbol interference channels. Equalization is performed using the minimum mean squared error criterion. Three linear equalizers are proposed. The exact equalizer, a spatially and temporally varying equalizer, has performance similar to schemes based on message-passing and has lower complexity. To save further on computational cost, time-invariant implementations of the exact equalizer are also proposed. Extrinsic information transfer charts are used to analyze the equalizers. A density evolution algorithm using a Gaussian approximation is proposed and used to study the trade-off between computational cost and performance for different decoding schedules.