Search

Paddy is the main nourishment in the south sections of India. It straightforwardly influences the country's infrastructure. Bacterial leaf blight, brown spot and leaf smut were its main illnesses found in paddy field crops, which also significantly affect its profitability. This work presents an image processing system based automatic identification and classification of various paddy leaf diseases by affecting the cultivation of paddy namely bacterial leaf blight, brown spot, and leaf smut. The key phases of proposed methodology are pre-processing of images, segmentation, feature extraction and classification. The technique used in the proposed work to identify paddy leaf disease involves Otsu threshold method for segmentation, grey level co-occurrence metrics for feature extraction and k-nearest neighbors algorithm for classification. The proposed system aims at achieving rapid and accurate identification of disease and classification of the disease type and also helps in classifying the various paddy leaf diseases by utilizing the texture and shape features that majorly contributes in leaf disease recognition.

The problem of inferring the hidden state of individual nodes in social/sensor networks in which node activities affect their neighbors is growing in importance. We present an undirected generative model, a type of probabilistic model that has so far not been used for modeling latent variables influenced by neighbors in a network. We also propose an efficient inference method based on variational inference principles which, in contrast to sampling methods used in most existing models, is scalable to larger networks. While training is intractable in general, by using stochastic methods to approximate the intractable derivative, we show that our model can be trained using the maximum likelihood method by formulating the model as an exponential family distribution. The results demonstrate that the proposed undirected model can accurately infer latent states compared to baseline methods.

We study convergence of iterated belief revision in complex fusion environments, which may consist of a network of soft (i.e., human or human-based) and hard (i.e., conventional physics-based) sensors and where agent communications may be asynchronous and the link structure may be dynamic. In particular, we study the problem in which network agents exchange and revise belief functions (which generalize probability mass functions) and are more geared towards handling the uncertainty pervasive in soft/hard fusion environments. We focus on belief revision in which agents utilize a generalized fusion rule that is capable of generating a rational consensus. It includes the widely used weighted average consensus as a special case. By establishing this fusion scheme as a pool of paracontracting operators, we derive general convergence criteria that are relevant for a wide range of applications. Furthermore, we analyze the conditions for consensus for various social networks by simulating several network topologies and communication patterns that are characteristic of such networks.

In linear prediction of speech, the 1-norm error minimization criterion has been shown to provide a valid alternative to the 2-norm minimization criterion. However, unlike 2-norm minimization, 1-norm minimization does not guarantee the stability of the corresponding all-pole filter and can generate saturations when this is used to synthesize speech. In this paper, we introduce two new methods to obtain intrinsically stable predictors with the 1-norm minimization. The first method is based on constraining the roots of the predictor to lie within the unit circle by reducing the numerical range of the shift operator associated with the particular prediction problem considered. The second method uses the alternative Cauchy bound to impose a convex constraint on the predictor in the 1-norm error minimization. These methods are compared with two existing methods: the Burg method, based on the 1-norm minimization of the forward and backward prediction error, and the iteratively reweighted 2-norm minimization known to converge to the 1-norm minimization with an appropriate selection of weights. The evaluation gives proof of the effectiveness of the new methods, performing as well as unconstrained 1-norm based linear prediction for modeling and coding of speech.

Dempster-Shafer (DS) belief theory provides a convenient framework for the development of powerful data fusion engines by allowing for a convenient representation of a wide variety of data imperfections. The recent work on the DS theoretic (DST) conditional approach, which is based on the Fagin-Halpern (FH) DST conditionals, appears to demonstrate the suitability of DS theory for incorporating both soft (generated by human-based sensors) and hard (generated by physics-based sources) evidence into the fusion process. However, the computation of the FH conditionals imposes a significant computational burden. One reason for this is the difficulty in identifying the FH conditional core, i.e., the set of propositions receiving nonzero support after conditioning. The conditional core theorem (CCT) in this paper redresses this shortcoming by explicitly identifying the conditional focal elements with no recourse to numerical computations, thereby providing a complete characterization of the conditional core. In addition, we derive explicit results to identify those conditioning propositions that may have generated a given conditional core. This "converse" to the CCT is of significant practical value for studying the sensitivity of the updated knowledge base with respect to the evidence received. Based on the CCT, we also develop an algorithm to efficiently compute the conditional masses (generated by FH conditionals), provide bounds on its computational complexity, and employ extensive simulations to analyze its behavior.

Gaussian mixture model (GMM)-based predictive coding of line spectral frequencies (LSFs) has gained wide acceptance. In such coders, each mixture of a GMM can be interpreted as defining a linear predictive transform coder. In this paper, we use Kalman filtering principles to model each of these linear predictive transform coders to present GMM Kalman predictive coding. In particular, we show how suitable modeling of quantization noise leads to an adaptive a posteriori GMM that defines a signal-adaptive predictive coder that provides improved coding of LSFs in comparison with the baseline recursive GMM predictive coder. Moreover, we show how running the GMM Kalman predictive coders to convergence can be used to design a stationary GMM Kalman predictive coding system which again provides improved coding of LSFs but now with only a modest increase in run-time complexity over the baseline. In packet loss conditions, this stationary GMM Kalman predictive coder provides much better performance than the recursive GMM predictive coder, and in fact has comparable mean performance to a memoryless GMM coder. Finally, we illustrate how one can utilize Kalman filtering principles to design a postfilter which enhances decoded vectors from a recursive GMM predictive coder without any modifications to the encoding process.

All-pole spectral envelope estimates based on linear prediction (LP) for speech signals often exhibit unnaturally sharp peaks, especially for high-pitch speakers. In this paper, regularization is used to penalize rapid changes in the spectral envelope, which improves the spectral envelope estimate. Based on extensive experimental evidence, we conclude that regularized linear prediction outperforms bandwidth-expanded linear prediction. The regularization approach gives lower spectral distortion on average, and fewer outliers, while maintaining a very low computational complexity.

We present all-pole models based upon the minimum variance distortionless response (MVDR) spectrum for spectral modeling of speech. The MVDR method, which is popular in array processing, provides all-pole spectra that are robust for modeling both voiced and unvoiced speech. Although linear prediction (LP) is a popular method for obtaining all-pole model parameters, LP spectral envelopes overestimate and overemphasize the medium and high pitch voiced speech spectral powers, thereby featuring unwanted sharp contours, and do not improve in spectral envelope modeling performance as the filter order is increased. In contrast, the MVDR all-pole spectrum which can be easily obtained from the LP coefficients, features improved spectral envelope modeling as the filter order is increased. In particular, the high order MVDR spectrum models voiced speech spectra very well, particularly at the perceptually important harmonics, and features a smooth contoured envelope. Furthermore, the MVDR spectrum can be based upon either conventional time domain correlation estimates or upon spectral samples, a task that is common in frequency domain speech coding. In particular, the MVDR spectrum of sufficient order provides an all-pole envelope that models a set of spectral samples exactly. In addition, the MVDR all-pole spectrum is also suitable for modeling unvoiced speech spectra.

In this paper, we develop a method to find the uncertain consequent by fusing the uncertain antecedent and the uncertain implication rule. In particular with Dempster-Shafer theoretic models utilized to capture the uncertainty intervals associated with the antecedent and the rule itself, we derive bounds on the confidence interval associated with the rule consequent. We derive inequalities for the belief and plausibility values of the consequent and with least commitment choice they become equations. We also demonstrate the consistency of our model with probability and classical logic.

Quantization of signals is required for many transmission, storage and compression applications. The original signal is quantized at the encoder side. At the decoder side, a replica of the original signal that should resemble the original signal in some sense is recovered. Present quantizers make an effort to reduce the distortion of the signal in the sense of reproduction fidelity. Consider scenarios in which signals are generated from multiple classes. The encoder focuses on the task of quantizing the data without any regards to the class of the signal. The quantized signal reaches the decoder where not only the recovery of the signal should take place but also a decision is to be made on the class of the signal based on the quantized version of the signal only. In this paper, we study the design of such scalar quantizer that is optimized for the task of classification at the decoder. We define the distortion to be the symmetric Kullback-Leibler (KL) divergence measure between the conditional probabilities of class given the signal before and after quantization. A high-rate analysis of the quantizer is presented and the optimum point density of the quantizer for minimizing the symmetric KL divergence is derived. The performance of this method on synthetically generated data is examined and observed to be superior in the task of classification of signals at the decoder.

Incorporation of soft evidence into the fusion process poses considerable challenges, including issues related to the material implications of propositional logic statements, contradictory evidence, and non-identical scopes of sources providing soft evidence. The conditional approach to Dempster-Shafer (DS) theoretic evidence updating and fusion provides a promising avenue for overcoming these challenges. However, the computation of the Fagin-Halpern (FH) conditionals utilized in the conditional evidence updating strategies is non-trivial because of the lack of a method to identify the conditional focal elements directly. The work in this paper presents a complete characterization of the conditional focal elements via a necessary and sufficient condition that identifies the explicit structure of a proposition that will remain a focal element after conditioning. We illustrate the resulting computational advantage via several experiments.

Linear prediction of speech based on 1-norm minimization has already proved to be an interesting alternative to 2-norm minimization. In particular, choosing the 1-norm as a convex relaxation of the 0-norm, the corresponding linear prediction model offers a sparser residual better suited for coding applications. In this paper, we propose a new speech modeling technique based on reweighted 1-norm minimization. The purpose of the reweighted scheme is to overcome the mismatch between 0-norm minimization and 1-norm minimization while keeping the problem solvable with convex estimation tools. Experimental results prove the effectiveness of the reweighted 1-norm minimization, offering better coding properties compared to 1-norm minimization.

Gaussian Mixture Model (GMM)-based vector quantization of Line Spectral Frequencies (LSFs) has gained wide acceptance in speech coding. In predictive coding of LSFs, the GMM approach utilizing Kalman filtering principles to account for quantization noise has been shown to perform better than a baseline GMM Recursive Coder approaches for both clean and packet loss conditions at roughly the same complexity. However, the GMM Kalman based predictive coder was not specifically designed for operation in packet loss conditions. In this paper, we show how an initial GMM Kalman predictive coder can be utilized to obtain a robust GMM predictive coder specifically designed to operate in packet loss. In particular, we demonstrate how one can define sets of encoding and decoding modes, and design special Kalman encoding and decoding gains for each set. With this framework, GMM predictive coding design can be viewed as determining the special Kalman gains that minimize the expected minimum mean squared error at the decoder in packet loss conditions. The simulation results demonstrate that the proposed robust Kalman predictive coder achieves better performance than the baseline GMM predictive coders.

In this work, we propose a novel scheme to re-estimate the linear predictive parameters in sparse speech coding. The idea is to estimate the optimal truncated impulse response that creates the given sparse coded residual without distortion. An all-pole approximation of this impulse response is then found using a least square approximation. The all-pole approximation is a stable linear predictor that allows a more efficient reconstruction of the segment of speech. The effectiveness of the algorithm is proved in the experimental analysis.

Gaussian mixture model (GMM)-based predictive coding of line spectral frequencies (LSFs) has gained wide acceptance. In such coders, each mixture of a GMM can be interpreted as defining a linear predictive transform coder. In this paper we optimize each of these linear predictive transform coders using Kalman predictive coding techniques to present GMM Kalman predictive coding. In particular, we show how suitable modeling of quantization noise leads to an adaptive a-posteriori GMM that defines a signal-adaptive predictive coder that provides superior coding of LSFs in comparison with the baseline GMM predictive coder. Moreover, we show how running the Kalman predictive coders to convergence can be used to design a stationary predictive coding system which again provides superior coding of LSFs but now with no increase in run-time complexity over the baseline.

To overcome the limitations of linear prediction based all-pole models of voiced speech, we present some new methods for obtaining the parameters of all-pole filters. In particular, we present a technique based on the minimum variance distortionless response (MVDR) spectrum estimation method. With a sufficient filter order, MVDR filters model voiced speech formants exactly. Utilizing a property of MVDR all-pole filters, we present a technique for obtaining low order perceptually weighted all-pole filters. These new all-pole models provide superior modeling of medium pitch and high pitch voiced speech.

In this paper we propose the MVDR method, which is based upon the minimum variance distortionless response (MVDR) spectrum estimation method, for modeling voiced speech. Developed to overcome some of the shortcomings of linear prediction models, the MVDR method provides better models for medium and high pitch voiced speech. The MVDR model is an all-pole model whose spectrum is easily obtained from a modest non-iterative computation involving the linear prediction coefficients thereby retaining some of the computational attractiveness of LPC methods. With the proper choice of filter order, which is dependent on the number of harmonics, the MVDR spectrum models the formants and spectral powers of voiced speech exactly. An efficient reduced model order MVDR method is developed to further enhance its applicability. An extension of the reduced order MVDR method for recovering the correct amplitudes of the harmonics of voiced speech is also presented.

In this paper three methods of basis selection are considered: basic matching pursuit (BMP), order recursive matching pursuit (ORMP) and modified matching pursuit (MMP). These algorithms are briefly described and particular attention is paid, in the formulation of these algorithms, to the computation required. Fast versions of the algorithms are developed. The algorithms are evaluated in terms of their ability to produce a sparse solution and also in terms of their computational complexity and the storage necessary to implement them. Complexity-wise, BMP and MMP are shown to be comparable while ORMP is the most complex. In terms of their ability to select basis vectors, ORMP was the best followed by MMP and then BMP.

We present several analytical properties of minimum variance distortionless response (MVDR) based all-pole models that demonstrate the advantages and usefulness of these models for speech spectral coding. In particular, we show that a sufficient order MVDR all-pole model provides a spectral envelope that fits a set of spectral samples exactly with a parameterization convenient for quantization purposes. In addition, we show that MVDR all-pole filters provide a monotonically decreasing spectral distortion with increasing filter order. Furthermore, we show that the MVDR all-pole filter possesses the flexibility to be obtained from correlations based upon either spectral samples or conventional time-domain correlations. Finally, exploiting the insight gained from MVDR modeling, we introduce a novel class of constrained all-pole models for efficient spectral coding. In this approach, a subset of the line spectral frequency (LSF) parameters associated with the all-pole model are judiciously fixed, leading to a simpler model parameterization.

Let R = (rij) be an m × n matrix and let S = (sik ) be dip × q matrix denned over a field F. The Kronecker product R × S of R and S is denned as follows: Definition 1.1. The Kronecker product R × S of the matrices R and S is given by where rij S; i = 1, 2, … , m; j = 1, 2, … , n, is itself a p × q matrix (1, 69-70).