Separable Input Research Articles

Gesture recognition framework for upper-limb prosthetics using entropy features from electromyographic signals and a Gaussian kernel SVM classifier

This paper puts forward a novel entropy features based multi-class SVM classifier framework to predict the limb movement of the transradial amputees from the surface electromyography (sEMG) signals. The major challenges with the sEMG signal are nonlinear and non-stationary characteristics and susceptibility to noise. Consequently, a robust and an effective feature extraction framework which is invariant to force level variations is central in sEMG based prosthesis control. To address the aforementioned challenges, this study leverages the potential of variational mode decomposition (VMD) technique to identify the prominent frequency modes of the sEMG signals, and performs the spectral evaluation of the decomposed sEMG modes to identify the dominant ones to extract the entropy features. Subsequently, we evaluate the efficacy of four nonlinear optimal feature selection techniques and identify the prominent entropy features to train the multi-class SVM model that can predict the gestures. Specifically, to handle the nonlinearly separable input data, this study implements a kernelization named a radial basis function (RBF), which has good generalization and noise tolerance features. The efficacy of the proposed framework is tested using the publicly available datasets that contain gestures from transradial and congenital amputees for functional gestures. Experimental results obtained for various gestures with dynamic force levels underscore that the proposed framework is highly robust against the force level variations and can achieve a classification accuracy of 99.07%.

Hybrid genetic optimization for quantum feature map design

Kernel methods are an import class of techniques in machine learning. To be effective, good feature maps are crucial for mapping non-linearly separable input data into a higher dimensional (feature) space, thus allowing the data to be linearly separable in feature space. Previous work has shown that quantum feature map design can be automated for a given dataset using NSGA-II, a genetic algorithm, while both minimizing circuit size and maximizing classification accuracy. However, the evaluation of the accuracy achieved by a candidate feature map is costly. In this work, we demonstrate the suitability of kernel-target alignment as a substitute for accuracy in genetic algorithm-based quantum feature map design. Kernel-target alignment is faster to evaluate than accuracy and does not require some data points to be reserved for its evaluation. To further accelerate the evaluation of genetic fitness, we provide a method to approximate kernel-target alignment. To improve kernel-target alignment and root mean squared error, the final trainable parameters of the generated circuits are further trained using COBYLA to determine whether a hybrid approach applying conventional circuit parameter training can easily complement the genetic structure optimization approach. A total of eight new approaches are compared to the original across nine varied binary classification problems from the UCI machine learning repository, showing that kernel-target alignment and its approximation produce feature map circuits enabling comparable accuracy to the previous work but with larger margins on training data (in excess of 20% larger) that improve further with circuit parameter training.

The role of input vs. output phonological working memory in narrative production: Evidence from case series and case study approaches

ABSTRACT Separable input and output phonological working memory (WM) capacities have been proposed, with the input capacity supporting speech recognition and the output capacity supporting production. We examined the role of input vs. output phonological WM in narrative production, examining speech rate and pronoun ratio – two measures with prior evidence of a relation to phonological WM. For speech rate, a case series approach with individuals with aphasia found no significant independent contribution of input or output phonological WM capacity after controlling for single-word production. For pronoun ratio, there was some suggestion of a role for input phonological WM. Thus, neither finding supported a specific role for an output phonological buffer in speech production. In contrast, two cases demonstrating dissociations between input and output phonological WM capacities provided suggestive evidence of predicted differences in narrative production, though follow-up research is needed. Implications for case series vs. case study approaches are discussed.

Exact dynamics of multimode periodic input states in coupled waveguide arrays

In this work, we investigate the quantum state reconstruction of the periodic input state in a 1-D waveguide array. In particular, we consider a single-photon multimode entangled W-state with different periodicities as an input to the array and study the effect of periodicity in the output. For comparison, we also study separable single photon periodic input states. We study the evolution of average photon number and the fidelity of the periodic input state and give the exact solution to investigate the revival of single photon multimode periodic input. Our solution is valid for any number of waveguides in the array. Our findings show the revival of the average photon number and almost complete quantum state reconstruction of the initial multimode entangled state for certain periodicities. The results reported here are significant because multimode-entangled states are essential resources for various applications in the physical implementation of photonic quantum technologies.

Work Extraction Processes from Noisy Quantum Batteries: The Role of Nonlocal Resources.

We demonstrate an asymmetry between the beneficial effects one can obtain using nonlocal operations and nonlocal states to mitigate the detrimental effects of environmental noise in the work extraction process from quantum battery models. Specifically, we show that using nonlocal recovery operations after the noise action can, in general, increase the amount of work one can recover from the battery even with separable (i.e., nonentangled) input states. On the contrary, employing entangled input states with local recovery operations will generally not improve the battery performance.

The Role of Fidelity in Goal-Oriented Semantic Communication: A Rate Distortion Approach

We study a variant of a robust description source coding framework, which is a relevant model for goal-oriented semantic information transmission, via its corresponding characterization. Considering two individual single-letter separable distortion constraints and input and output data acting as the intrinsic and extrinsic message, respectively, we first derive a lower bound on the optimal rates of the problem, as well as necessary and sufficient conditions for this bound to be tight. Subsequently, we prove a general result that provides in parametric form the optimal solution of the characterization of this problem. Capitalizing on these results, we examine the structure of the solution for one case study of general binary alphabets under Hamming distortions and solve in closed form a special case. We also solve another general binary alphabet case where a Hamming and an erasure distortion coexist, as a means to highlight the importance of selecting the type of the distortion constraint in goal-oriented semantic communication. Furthermore, we develop a goal-oriented Blahut-Arimoto (BA) algorithm, which can be used for the computation of any finite alphabet intrinsic or extrinsic message under individual distortion criteria. Finally, we revisit the problem for multidimensional independent and identically distributed (i.i.d.) jointly Gaussian processes with individual mean-square error (MSE) distortion constraints, providing new insights that have previously been overlooked. This work reveals the cardinal role of context-dependent fidelity criteria in goal-oriented semantic communication.

Activation functions in deep learning: A comprehensive survey and benchmark

Neural networks have shown tremendous growth in recent years to solve numerous problems. Various types of neural networks have been introduced to deal with different types of problems. However, the main goal of any neural network is to transform the non-linearly separable input data into more linearly separable abstract features using a hierarchy of layers. These layers are combinations of linear and nonlinear functions. The most popular and common non-linearity layers are activation functions (AFs), such as Logistic Sigmoid, Tanh, ReLU, ELU, Swish and Mish. In this paper, a comprehensive overview and survey is presented for AFs in neural networks for deep learning. Different classes of AFs such as Logistic Sigmoid and Tanh based, ReLU based, ELU based, and Learning based are covered. Several characteristics of AFs such as output range, monotonicity, and smoothness are also pointed out. A performance comparison is also performed among 18 state-of-the-art AFs with different networks on different types of data. The insights of AFs are presented to benefit the researchers for doing further research and practitioners to select among different choices. The code used for experimental comparison is released at: https://github.com/shivram1987/ActivationFunctions.

Joint optic disc and cup segmentation based on densely connected depthwise separable convolution deep network

BackgroundGlaucoma is an eye disease that causes vision loss and even blindness. The cup to disc ratio (CDR) is an important indicator for glaucoma screening and diagnosis. Accurate segmentation for the optic disc and cup helps obtain CDR. Although many deep learning-based methods have been proposed to segment the disc and cup for fundus image, achieving highly accurate segmentation performance is still a great challenge due to the heavy overlap between the optic disc and cup.MethodsIn this paper, we propose a two-stage method where the optic disc is firstly located and then the optic disc and cup are segmented jointly according to the interesting areas. Also, we consider the joint optic disc and cup segmentation task as a multi-category semantic segmentation task for which a deep learning-based model named DDSC-Net (densely connected depthwise separable convolution network) is proposed. Specifically, we employ depthwise separable convolutional layer and image pyramid input to form a deeper and wider network to improve segmentation performance. Finally, we evaluate our method on two publicly available datasets, Drishti-GS and REFUGE dataset.ResultsThe experiment results show that the proposed method outperforms state-of-the-art methods, such as pOSAL, GL-Net, M-Net and Stack-U-Net in terms of disc coefficients, with the scores of 0.9780 (optic disc) and 0.9123 (optic cup) on the DRISHTI-GS dataset, and the scores of 0.9601 (optic disc) and 0.8903 (optic cup) on the REFUGE dataset. Particularly, in the more challenging optic cup segmentation task, our method outperforms GL-Net by 0.7% in terms of disc coefficients on the Drishti-GS dataset and outperforms pOSAL by 0.79% on the REFUGE dataset, respectively.ConclusionsThe promising segmentation performances reveal that our method has the potential in assisting the screening and diagnosis of glaucoma.

From squeezing to Gaussian entanglement via beamsplitters

We investigate relations between squeezing of separable input states and entanglement of output states generated by a beamsplitter, both of which are valuable nonclassical resources for quantum information processing. Motivated by the seminal work of Asboth´ et al [2005 Phys. Rev. Lett. 94 173602], where a systematic study of nonclassicality (including squeezing) and beamsplitter-generated entanglement is conducted, we consider the situation when the two input states are all Gaussian and identify certain constraints on the squeezing parameters to generate entanglement. Contrary to naive expectation, large squeezing does not necessarily guarantee large entanglement. Depending on the phase relations between the input states and the beamsplitters, the conditions to generate entanglement vary, which may be regarded as a manifestation of interference. In two typical cases, it is the difference or sum of the squeezing amplitudes that plays a key role in generating entanglement. Analytic expressions of entanglement in terms of the difference or sum of the squeezing amplitudes are obtained, which are further shown to be monotonic in the relevant parameters.

Functional characterization of retinal ganglion cells using tailored nonlinear modeling

The mammalian retina encodes the visual world in action potentials generated by 20–50 functionally and anatomically-distinct types of retinal ganglion cell (RGC). Individual RGC types receive synaptic input from distinct presynaptic circuits; therefore, their responsiveness to specific features in the visual scene arises from the information encoded in synaptic input and shaped by postsynaptic signal integration and spike generation. Unfortunately, there is a dearth of tools for characterizing the computations reflected in RGC spike output. Therefore, we developed a statistical model, the separable Nonlinear Input Model, to characterize the excitatory and suppressive components of RGC receptive fields. We recorded RGC responses to a correlated noise (“cloud”) stimulus in an in vitro preparation of mouse retina and found that our model accurately predicted RGC responses at high spatiotemporal resolution. It identified multiple receptive fields reflecting the main excitatory and suppressive components of the response of each neuron. Significantly, our model accurately identified ON-OFF cells and distinguished their distinct ON and OFF receptive fields, and it demonstrated a diversity of suppressive receptive fields in the RGC population. In total, our method offers a rich description of RGC computation and sets a foundation for relating it to retinal circuitry.

Analysis of Model and Iteration Dependencies in Distributed Feasible-Point Algorithms for Optimal Control Computation

The problem of computing optimal control inputs is studied for networks of dynamical linear systems, with respect to separable input constraints and a separable quadratic cost over a finite time-horizon. The main results concern network structures for which the iterates of three feasible-point algorithms can be computed exactly on a subsystem-by-subsystem basis with access restricted to local model-data and algorithm-state information. In particular, hop-based network proximity bounds are investigated for algorithms based on projected gradient, random co-ordinate descent and Jacobi iterations, via graph-based characterisations of various aspects of an equivalent static formulation of the optimal control problem.

Tailoring spin chain dynamics for fractional revivals

The production of quantum states required for use in quantum protocols & technologies is studied by developing the tools to re-engineer a perfect state transfer spin chain so that a separable input excitation is output over multiple sites. We concentrate in particular on cases where the excitation is superposed over a small subset of the qubits on the spin chain, known as fractional revivals, demonstrating that spin chains are capable of producing a far greater range of fractional revivals than previously known, at high speed. We also provide a numerical technique for generating chains that produce arbitrary single-excitation states, such as the W state.

Dynamical analysis of quantum linear systems driven by multi-channel multi-photon states

In this paper, we investigate the dynamics of quantum linear systems where the input signals are multi-channel multi-photon states, namely states determined by a definite number of photons superposed in multiple input channels. In contrast to most existing studies on separable input states in the literature, we allow the existence of quantum correlation (for example quantum entanglement) in these multi-channel multi-photon input states. Due to the prevalence of quantum correlations in the quantum regime, the results presented in this paper are very general. Moreover, the multi-channel multi-photon states studied here are reasonably mathematically tractable. Three types of multi-photon states are considered: (1) m photons superposed among m channels, (2) N photons superposed among m channels where N≥m, and (3) N photons superposed among m channels where N is an arbitrary positive integer. Formulae for intensities and states of output fields are derived. Examples are used to demonstrate the effectiveness of the results.

Authorship Authentication for Twitter Messages Using Support Vector Machine

With the rapid growth of internet usage, authorship authentication of online messages became challenging research topic in the last decades. In this paper, we used a team of support vector machines to authenticate 5 Twitter authors’ messages. SVM is one of the commonly used and strong classification algorithms in authorship attribution problems. SVM maps the linearly non separable input data to a higher dimensional space by a hyperplane via radial base functions. Firstly using the training data, 10 hyperplanes that separate pair wise five authors training data are built. Then the expertise of these SVMs combined to classify the testing data into five classes. 20 tweets with 16 features from each author were used for evaluation. In spite of the randomly choice of the features, one of the author accuracy around 75% is achieved.

Tomographic characterization of a linear optical quantum Toffoli gate

We report on a detailed characterization of a three-qubit linear optical quantum Toffoli gate. Our experiment utilizes correlated photon pairs generated in the process of spontaneous parametric down-conversion. Two qubits are encoded into polarization and spatial degrees of freedom of a signal photon, and the third qubit is represented by polarization of an idler photon. The linear optical Toffoli gate is implemented by interference of photons on a partially polarizing beam splitter inserted inside a Mach Zehnder interferometer formed by two calcite beam displacers. We have measured 4032 different two-photon coincidences, which allows us to estimate the fidelity of the gate to be 90%. Although these data are not tomographically complete, we show that they are sufficient for a reliable reconstruction of the quantum process matrix of the gate via the recently proposed maximum likelihood--maximum entropy estimation procedure. To probe the entangling capability of the gate, we have investigated generation of three-qubit GHZ states from fully and partially separable input states and we have performed a full tomography of the output states. We compare the reconstructed states with theoretical predictions obtained with the use of the estimated quantum process matrix and obtain a very good agreement.

The identification of neuro-fuzzy based MIMO Hammerstein model with separable input signals

A novel identification method of neuro-fuzzy based MIMO Hammerstein model by using the correlation analysis method is presented in this paper. A special test signal that contains independent separable signals and uniformly random multi-step signal is adopted to identify the MIMO Hammerstein process, resulting in the identification problem of the linear model separated from that of nonlinear part. As a result, the identification of the dynamic linear element can be separated from the static nonlinear element without any redundant adjustable parameters. Moreover, it can circumvent the problem of initialization and convergence of the model parameters discussed in the existing iterative algorithms used for identification of MIMO Hammerstein model. Examples are used to illustrate the effectiveness of the proposed method.

Channel-capacity gain in entanglement-assisted communication protocols based exclusively on linear optics, single-photon inputs, and coincidence photon counting

Entanglement can effectively increase communication channel capacity as evidenced by dense coding that predicts a capacity gain of $1\phantom{\rule{4.pt}{0ex}}\text{bit}$ when compared to entanglement-free protocols. However, dense coding relies on Bell states and when implemented using photons the capacity gain is bounded by $0.585\phantom{\rule{4.pt}{0ex}}\text{bits}$ due to one's inability to discriminate between the four optically encoded Bell states. In this paper we study the following question: Are there alternative entanglement-assisted protocols that rely only on linear optics, coincidence photon counting, and separable single-photon input states and at the same time provide a greater capacity gain than $0.585\phantom{\rule{4.pt}{0ex}}\text{bits}$? We show that besides the Bell states there is a class of bipartite four-mode two-photon entangled states that facilitate an increase in channel capacity. We also discuss how the proposed scheme can be generalized to the case of two-photon $N$-mode entangled states for $N=6,8$.

Nonlocal quantum gate on quantum continuous variables with minimal resources

We experimentally demonstrate, with an all-optical setup, a nonlocal deterministic quantum nondemolition interaction gate applicable to quantum states at nodes separated by a physical distance and connected by classical communication channels. The gate implementation, based on entangled states shared in advance, local operations, and classical communication, runs completely in parallel fashion at both of the local nodes, requiring minimum resources. The nondemolition character of the gate up to the local unitary squeezing is verified by the analysis using several coherent states. A genuine quantum nature of the gate is confirmed by the capability of deterministically producing an entangled state at the output from two separable input states. The all-optical nonlocal gate operation can be potentially incorporated into distributed quantum computing with atomic or solid-state systems as a cross-processor unitary operation.

A quantum gate between a flying optical photon and a single trapped atom

The steady increase in control over individual quantum systems supports the promotion of a quantum technology that could provide functionalities beyond those of any classical device. Two particularly promising applications have been explored during the past decade: photon-based quantum communication, which guarantees unbreakable encryption but which still has to be scaled to high rates over large distances, and quantum computation, which will fundamentally enhance computability if it can be scaled to a large number of quantum bits (qubits). It was realized early on that a hybrid system of light qubits and matter qubits could solve the scalability problem of each field--that of communication by use of quantum repeaters, and that of computation by use of an optical interconnect between smaller quantum processors. To this end, the development of a robust two-qubit gate that allows the linking of distant computational nodes is "a pressing challenge". Here we demonstrate such a quantum gate between the spin state of a single trapped atom and the polarization state of an optical photon contained in a faint laser pulse. The gate mechanism presented is deterministic and robust, and is expected to be applicable to almost any matter qubit. It is based on reflection of the photonic qubit from a cavity that provides strong light-matter coupling. To demonstrate its versatility, we use the quantum gate to create atom-photon, atom-photon-photon and photon-photon entangled states from separable input states. We expect our experiment to enable various applications, including the generation of atomic and photonic cluster states and Schrödinger-cat states, deterministic photonic Bell-state measurements, scalable quantum computation and quantum communication using a redundant quantum parity code.

The Relationship between Language Production and Verbal Short-Term Memory: The Role of Stress Grouping

This study investigates the influence of stress grouping on verbal short-term memory (STM). English speakers show a preference to combine syllables into trochaic groups, both lexically and in continuous speech. In two serial recall experiments, auditory lists of nonsense syllables were presented with either trochaic (STRONG-weak) or iambic (weak-STRONG) stress patterns, or in monotone. The acoustic correlates that carry stress were also manipulated in order to examine the relationship between input and output processes during recall. In Experiment 1, stressed and unstressed syllables differed in intensity and pitch but were matched for spoken duration. Significantly more syllables were recalled in the trochaic stress pattern condition than in the iambic and monotone conditions, which did not differ. In Experiment 2, spoken duration and pitch were manipulated but intensity was held constant. No effects of stress grouping were observed, suggesting that intensity is a critical acoustic factor for trochaic grouping. Acoustic analyses demonstrated that speech output was not identical to the auditory input, but that participants generated correct stress patterns by manipulating acoustic correlates in the same way in both experiments. These data challenge the idea of a language-independent STM store and support the notion of separable phonological input and output processes.