Binary Input Research Articles

On the Compressive Power of Autoencoders With Linear and ReLU Activation Functions.

In this letter, we mainly study the depth and width of autoencoders consisting of rectified linear unit (ReLU) activation functions. An autoencoder is a layered neural network consisting of an encoder, which compresses an input vector to a lower-dimensional vector, and a decoder, which transforms the low-dimensional vector back to the original input vector exactly (or approximately). In a previous study, Melkman etal. (2023) studied the depth and width of autoencoders using linear threshold activation functions with binary input and output vectors. We show that similar theoretical results hold if autoencoders using ReLU activation functions with real input and output vectors are used. Furthermore, we show that it is possible to compress input vectors to one-dimensional vectors using ReLU activation functions, although the size of compressed vectors is trivially Ω(log n) for autoencoders with linear threshold activation functions, where n is the number of input vectors. We also study the cases of linear activation functions. The results suggest that the compressive power of autoencoders using linear activation functions is considerably limited compared with those using ReLU activation functions.

Scalable Multispecies Ion Transport in a Grid-Based Surface-Electrode Trap

Quantum processors based on linear arrays of trapped ions have achieved exceptional performance, but scaling to large qubit numbers requires realizing two-dimensional ion arrays as envisioned in the quantum charge-coupled device (QCCD) architecture. Here, we present a scalable method for the control of ion crystals in a grid-based surface-electrode Paul trap and characterize it in the context of transport operations that sort and reorder multispecies crystals. By combining cowiring of control electrodes at translationally symmetric locations in each grid site with the sitewise ability to exchange the voltages applied to two special electrodes gated by a binary input, site-dependent operations can be achieved using only a fixed number of analog voltage signals and a single digital input per site. In two separate experimental systems containing nominally identical grid traps, one using Yb+171−Ba+138 crystals and the other Ba+137−Sr+88, we demonstrate this method by characterizing the conditional intrasite crystal reorder and the conditional exchange of ions between adjacent sites on the grid. Averaged across a multisite region of interest, we measure subquanta motional excitation in the axial in-phase and out-of-phase modes of the crystals following these operations at exchange rates of 2.5 kHz. In this initial demonstration, the logic controlling the voltage exchange occurs in software, but the applied signals mimic a proposed hardware implementation using crossover switches. These techniques can be further extended to implement other conditional operations in the QCCD architecture such as gates, initialization, and measurement. Published by the American Physical Society 2024

Estimation of probabilities of transitions of markov binary input signal of nonlinear system

The problem of estimating unknown probabilities of transitions of a random Markov binary input signal of a nonlinear one-dimensional discrete system based on estimating the expectation and variance of the output signal is considered. The defined expressions are built on the basis of considering equally probable transitions and the steady-state mode of the algorithm for assessing the state of the system, obtained by approximating the probability density of its output signal by the Pearson type I distribution. An example of comparison of theoretical calculations with the results of imitation mathematical modeling is given.

Simulation and Controller Design for a Fish Robot with Control Fins.

In this paper, a nonlinear simulation block for a fish robot was designed using MATLAB Simulink. The simulation block incorporated added masses, hydrodynamic damping forces, restoring forces, and forces and moments due to dorsal fins, pectoral fins, and caudal fins into six-degree-of-freedom equations of motion. To obtain a linearized model, we used three different nominal surge velocities (i.e., 0.2 m/s, 0.4 m/s, and 0.6 m/s). After obtaining output responses by applying pseudo-random binary signal inputs to a nonlinear model, an identification tool was used to obtain approximated linear models between inputs and outputs. Utilizing the obtained linearized models, two-degree-of-freedom proportional, integral, and derivative controllers were designed, and their characteristics were analyzed. For the 0.4 m/s nominal surge velocity models, the gain margins and phase margins of the surge, pitch, and yaw controllers were infinity and 69 degrees, 26.3 dB and 85 degrees, and infinity and 69 degrees, respectively. The bandwidths of surge, pitch, and yaw control loops were determined to be 2.3 rad/s, 0.17 rad/s, and 2.0 rad/s, respectively. Similar characteristics were observed when controllers designed for linear models were applied to the nonlinear model. When step inputs were applied to the nonlinear model, the maximum overshoot and steady-state errors were very small. It was also found that the nonlinear plant with three different nominal surge velocities could be controlled by a single controller designed for a linear model with a nominal surge velocity of 0.4 m/s. Therefore, controllers designed using linear approximation models are expected to work well with an actual nonlinear model.

On Bounding the Behavior of Neurons

A neuron with binary inputs and a binary output represents a Boolean function. Our goal is to extract this Boolean function into a tractable representation that will facilitate the explanation and formal verification of a neuron’s behavior. Unfortunately, extracting a neuron’s Boolean function is in general an NP-hard problem. However, it was recently shown that prime implicants of this Boolean function can be enumerated efficiently, with only polynomial time delay. Building on this result, we first propose a best-first search algorithm that is able to incrementally tighten the inner and outer bounds of a neuron’s Boolean function. Second, we show that these bounds correspond to truncated prime-implicant covers of the Boolean function. Next, we show how these bounds can be propagated in an elementary class of neural networks. Finally, we provide case studies that highlight our ability to bound the behavior of neurons.

Estimation of Probabilities of Transitions of a Markov Binary Input Signal of a Nonlinear System

Estimation of Probabilities of Transitions of a Markov Binary Input Signal of a Nonlinear System

A methodology of designing and optimising a 4-bit absolute-value detector for the application of Brain-computer Interface

Digital integrated circuits perform logical operations through several modules composed of logic gates. As a common module in digital circuits, the absolute value detector is applied to compare the absolute value of two binary inputs and can be further used for data sorting and searching. In clinical medicine, the detector can receive biological data from the sensor and control the instrument on the operating table. A typical application is Brain-computer Interface (BCI). Therefore, the performance of detectors is significant to electronic devices like computers. The optimisation of detectors usually concentrates on their propagation delay and power consumption. But theoretical optimisation will inevitably use simple logic and have large fan-in, and it is difficult to be implemented in practice. Here we explore an alternative plan that is more convenient to be produced in the industry, which combines static complementary metal-oxide semiconductor (CMOS) technology and transmission gates. After gate sizing, the critical path calculates the propagation delay and energy based on the logical effort theory. Finally, the relationship between energy and delay is analysed by scaling the voltage supply and the minimum energy occurs when the delay grows to 1.5 times the minimum delay.

Performance Comparison of Different HTM-Spatial Pooler Algorithms Based on Information-Theoretic Measures

Hierarchical temporal memory (HTM) is a promising unsupervised machine-learning algorithm that models key principles of neocortical computation. One of the main components of HTM is the spatial pooler (SP), which encodes binary input streams into sparse distributed representations (SDRs). In this paper, we propose an information-theoretic framework for the performance comparison of HTM-spatial pooler (SP) algorithms, specifically, for quantifying the similarities and differences between sparse distributed representations in SP algorithms. We evaluate SP's standalone performance, as well as HTM's overall performance. Our comparison of various SP algorithms using Renyi mutual information, Renyi divergence, and Henze–Penrose divergence measures reveals that the SP algorithm with learning and a logarithmic boosting function yields the most effective and useful data representation. Moreover, the most effective SP algorithm leads to superior HTM results. In addition, we utilize our proposed framework to compare HTM with other state-of-the-art sequential learning algorithms. We illustrate that HTM exhibits superior adaptability to pattern changes over time than long short term memory (LSTM), gated recurrent unit (GRU) and online sequential extreme learning machine (OS-ELM) algorithms. This superiority is evident from the lower Renyi divergence of HTM (0.23) compared to LSTM6000 (0.33), LSTM3000 (0.38), GRU (0.41), and OS-ELM (0.49). HTM also achieved the highest Renyi mutual information value of 0.79, outperforming LSTM6000 (0.73), LSTM3000 (0.71), GRU (0.68), and OS-ELM (0.62). These findings not only confirm the numerous advantages of HTM over other sequential learning algorithm, but also demonstrate the effectiveness of our proposed information-theoretic approach as a powerful framework for comparing and evaluating various learning algorithms.

State complexity of binary coded regular languages

For the given non-unary input alphabet Σ, a maximal prefix code h mapping strings over Σ to binary strings, and an optimal deterministic finite automaton (DFA) A with n states recognizing a language L over Σ, we consider the problem of how many states we need for an automaton A′ that decides membership in h(L), the binary coded version of L. Namely, A′ accepts binary inputs belonging to h(L) and rejects binary inputs belonging to h(LC), where LC is the complement of L. The outcome on inputs that are not valid binary codes for any string in Σ⁎ can be arbitrary: A′ may accept, reject, or halt in a “don't care” state. We show that every optimal deterministic don't care finite automaton (dcDFA) A′ solving this promise problem uses at most (‖Σ‖−1)⋅n states but at least n states. In addition, we have established a kind of pseudo-isomorphism between such DFAs and dcDFAs. We also show that, for each non-unary input alphabet Σ, there exists a maximal binary prefix code h such that, for each n≥2 and for each N in range from n to (‖Σ‖−1)⋅n, there exists a language L over Σ such that the optimal DFA recognizing L uses exactly n states and every optimal dcDFA for solving the above promise problem uses exactly N states. Thus, we have the complete state hierarchy for deciding membership in the binary coded version of L, with no magic numbers in between the lower and upper bounds.

Multi-LRA: Multi logical residual architecture for spiking neural networks

Recently, it has been noticed that certain current state-of-the-art (SOTA) spiking neural networks incorporated non-spike data by use of residual connections and count coding. However, these architectures may not be well-suited for the neuromorphic chips optimized primarily for processing binary inputs and floating-point weights. On the other hand, some pure-spike structures, such as SEW-AND and SEW-IAND, have been found to exhibit spike vanishing and limited capacity. To overcome these shortcomings, we analyze the energy consumption and expressive ability of binary and integer inputs on neuromorphic chips, and then propose a new Multi Logical Residual Architecture (Multi-LRA), which eliminates non-spike computation on neuromorphic chips and addresses the issues of spike vanishing and limited capacity in SEW-AND and SEW-IAND. Furthermore, we prove that the upper bound of conditional entropy of Multi-LRA is higher than that of SEW-AND and logical-OR, which means that Multi-LRA has better capacity and may avoid the spike vanishing. Finally, experiments on CIFAR-10 have shown that Multi-LRA can significantly reduce energy consumption because of its pure-spike computation and logical operation, where Multi-LRA-ResNet50 can reach the SOTA accuracy 96.27% in just 4 time steps. In addition, the effectiveness of Multi-LRA has been validated on ImageNet, CIFAR10-DVS and DVS128 Gesture.

Text Localization and Enhancement of Mobile Camera based Complex Natural Bilingual Text Scene Images

In this paper, we proposed a text localization and enhancement of the Mobile camera based complex natural scene images using fixed coordinates for removal of unwanted non text regions and making it localized. The morphological erosion technique is applied for the smoothness of foreground text regions. Here, the experimental working nature of the proposed method is first localized color input image is converted into a gray level, then a median filter is applied for the removal of noises. After that, the input image is further converted into binary image. In addition to this, morphological dilation and erosion operations also applied for connecting the broken pixels. Further, the connected component segmentation technique is applied to extract the foreground text regions from the binary input image. The real time input samples are collected by using mobile camera from the various categories of natural scenes of monolingual/bilingual text such as Stone, Wall, Poster and Iron-based. A total of 140 real time samples of natural scene text images are considered for the experimental setup. The experimental results shown that 132 real time sample images are localized and then obtained the accuracy of 94.24%. Further, bilingual text regions are enhanced using hybrid approach and obtain the segmentation accuracy of 83.17%. The novelty of the paper is working on the real time data set and it contains a mixture of printed and handwritten natural scene text images.

FormatFuzzer : Effective Fuzzing of Binary File Formats

Effective fuzzing of programs that process structured binary inputs, such as multimedia files, is a challenging task, since those programs expect a very specific input format. Existing fuzzers, however, are mostly format-agnostic, which makes them versatile, but also ineffective when a specific format is required. We present FormatFuzzer , a generator for format-specific fuzzers . FormatFuzzer takes as input a binary template (a format specification used by the 010 Editor) and compiles it into C++ code that acts as parser, mutator, and highly efficient generator of inputs conforming to the rules of the language. The resulting format-specific fuzzer can be used as a standalone producer or mutator in black-box settings, where no guidance from the program is available. In addition, by providing mutable decision seeds, it can be easily integrated with arbitrary format-agnostic fuzzers such as AFL to make them format-aware. In our evaluation on complex formats such as MP4 or ZIP, FormatFuzzer showed to be a highly effective producer of valid inputs that also detected previously unknown memory errors in ffmpeg and timidity .

LogicDT: a procedure for identifying response-associated interactions between binary predictors

Interactions between predictors play an important role in many applications. Popular and successful tree-based supervised learning methods such as random forests or logic regression can incorporate interactions associated with the considered outcome without specifying which variables might interact. Nonetheless, these algorithms suffer from certain drawbacks such as limited interpretability of model predictions and difficulties with negligible marginal effects in the case of random forests or not being able to incorporate interactions with continuous variables, being restricted to additive structures between Boolean terms, and not directly considering conjunctions that reveal the interactions in the case of logic regression. We, therefore, propose a novel method called logic decision trees (logicDT) that is specifically tailored to binary input data and helps to overcome the drawbacks of existing methods. The main idea consists of considering sets of Boolean conjunctions, using these terms as input variables for decision trees, and searching for the best performing model. logicDT is also accompanied by a framework for estimating the importance of identified terms, i.e., input variables and interactions between input variables. This new method is compared to other popular statistical learning algorithms in simulations and real data applications. As these evaluations show, logicDT is able to yield high prediction performances while maintaining interpretability.

Neural oscillatory correlates of motor vigor: an magnetoencephalographic study

Previous studies have shown that the anticipation of reward enhances motor performance, which reduces movement time and increases velocity [1]. In our recent study [2], we observed that when participants are required to infer the changing probabilities of a reward in a dynamic and uncertain setting, heightened expectations are consistently associated with faster motor performance. The study showed that performance time sensitivity to prediction strength remained consistent among both young and older healthy adults, as well as those with Parkinson’s disease. While the effects of dynamic motor strength have been observed, the neurological processes involved remain to be determined [3]. The study examined the neural oscillatory connections to motor vigor in dynamic and unpredictable settings. We used magnetoencephalography (MEG) and individual structural magnetic resonance imaging (MRI) to record readings from 25 healthy human participants (18 females) during the execution of our newly developed reward-based motor decision-making task [2]. This study used a reversal learning paradigm with shifting stimulus-outcome relationships. Participants were required to deduce which of two stimuli was linked to a reward on each trial, and indicate their choice through one of two finger press sequences, each with a distinct auditory response. The task was conducted in an unstable context, leading to fluctuations in the probability of reward associated with each response over time. First, we examined decision-making behavior using the validated Hierarchical Gaussian Filter (HGF, [4]). The model that most accurately described the behavioral data was the three-level “extended” HGF for binary categorical inputs, which is paired with a response model where decisions are dependent on the trial-wise estimate of volatility. This study allowed for the generation of reward probability trajectories on a trial-by-trial basis. Subsequently, applying Bayesian linear mixed models, we found a relationship between belief strength regarding reward contingencies and performance tempo on a trial-by-trial basis. The analysis of MEG signals is centered on reconstructing oscillatory activity sources using Linearly Constrained Minimum Variance beamforming [5]. Currently, we use convolution models in the source space to identify neural oscillatory correlations that differentiate motor performance and decision making. Next, we will evaluate connectivity patterns between frontal and motor regions that underlie the effects of motor invigoration. Identifying particular patterns of oscillatory connectivity that modulate motor vigor can provide insights into motor deficits observed in neurological and neuropsychiatric conditions associated with behavioral apathy.

Analysis Of 4-Bit Absolute Value Detector for ECG Signal Comparation

In the present era, where numerous heart conditions are prevalent, the significance of monitoring cardiac electrical signals has surged within the medical realm. Within this study, we leverage Complementary Metal Oxide Semiconductor (CMOS) and Pass-transistor Logic (PTL) methodologies for crafting and refining a 4-bit absolute value detection mechanism. This mechanism serves to identify and juxtapose electrocardiogram (ECG) signals. The 4-bit absolute value detection system is bifurcated into two core segments: one for absolute value identification and the other for comparative analysis. This setup facilitates a binary input's comparison with a predetermined threshold, thereby yielding a corresponding comparative outcome. Subsequently, a meticulous assessment is conducted on the lengthiest pathway, which is then subject to refinement through gate sizing and voltage supply (Vdd) adjustments. The findings unveil a delay of 34.13tp0 and an energy dissipation of 184.832 C. The resultant 4-bit absolute value detection mechanism, borne from this research institution's efforts, emerges as an invaluable asset in the medical domain, attributed to its exceptional optimization with respect to minimal delay and energy outlay.

Super-resolution flow-field reconstruction in rotating detonation combustors

The high-resolution flow field information is crucial for understanding the operational mechanisms of rotating detonation combustors (RDCs). However, due to limitations in sensor technology and observation techniques, the present experimental methods face great challenges in obtaining high-resolution data of RDCs. In the study, a novel physics-informed deep generative model is proposed for super-resolution flow-field reconstruction with low-resolution inputs. Compare to other methods, the application of physics-based constraints in this method enables accurate reconstruction of critical phenomena in RDCs while maintaining physics consistency. Additionally, the proposed super-resolution reconstruction method can be further employed for low sampling depth even binary input data, thereby evidently reducing the sampling depth per measurement point, which helps to overcome the challenges associated with high-frequency data acquisition and facility the implementation of larger-scale measurement arrays. As an extension to the measurement capabilities in reactive flow systems, the presented model possesses notable engineering significance.

Fast Grayscale Morphology for Circular Window

AbstractMorphological operations are among the most popular classic image filters. The filter assumes the maximum or minimum value within a window and is often used for light object thickening and thinning operations, which are important components of various workflows, such as object recognition and stylization. Circular windows are preferred over rectangular windows for obtaining isotropic filter results. However, the existing efficient algorithms focus on rectangular or binary input images. Efficient morphological operations with circular windows for grayscale images remain challenging. In this study, we present a fast grayscale morphology heuristic computation algorithm that decomposes circular windows using the convex hull of circles. We significantly accelerate traditional methods based on Minkowski addition by introducing new decomposition rules specialized for circular windows. As our morphological operation using a convex hull can be computed independently for each pixel, the algorithm is efficient for modern multithreaded hardware.

FPGA Based Efficient OFDM Based Design and Implementation for Data and Image Transmission for Healthcare

Enormous growth in telecommunication industry demands for high speed data transmission with better quality of service (Qos). The telecommunication networks are offering the services which is from 1 Mbps to several Mbps of speed. However, most of the existing techniques address to assure the very high speed data for multimedia communication. The multimedia data may find a suitable application in healthcare system. The OFDM modulation technique promises to provide the multimedia services at rather high speed using the spectrum more efficient compared traditional scheme like TDMA, FDMA. The orthogonality of carriers eliminates the interference among the closely packed carriers and offers comparatively efficient bandwidth. The OFDM design requires choosing proper parameter selection. Important feature of OFDM is that the multipaths are effectively eliminated by choosing a higher cyclic prefix values which, gives significant results but causing more energy loss. This paper presents an efficient design for OFDMtransceiver by using FPGA. The design is modeled and simulated using Matlab Simulink and finally the design is coded using Verilog RTL and simulated in modelsim and synthesizing and implementation is done using in Xilinx EDA tool. The image type of data is taken for transmission in the proposed OFDM transceiver system. The received image type data achieves PSNR value of 29.920dB and the binary input data achieves 36.06% improvement in power utilization and less area overhead. The paper also shows the improvement in area and power compared to existing authors.

Multimodal Mutual Information Maximization: A Novel Approach for Unsupervised Deep Cross-Modal Hashing.

In this article, we adopt the maximizing mutual information (MI) approach to tackle the problem of unsupervised learning of binary hash codes for efficient cross-modal retrieval. We proposed a novel method, dubbed cross-modal info-max hashing (CMIMH). First, to learn informative representations that can preserve both intramodal and intermodal similarities, we leverage the recent advances in estimating variational lower bound of MI to maximizing the MI between the binary representations and input features and between binary representations of different modalities. By jointly maximizing these MIs under the assumption that the binary representations are modeled by multivariate Bernoulli distributions, we can learn binary representations, which can preserve both intramodal and intermodal similarities, effectively in a mini-batch manner with gradient descent. Furthermore, we find out that trying to minimize the modality gap by learning similar binary representations for the same instance from different modalities could result in less informative representations. Hence, balancing between reducing the modality gap and losing modality-private information is important for the cross-modal retrieval tasks. Quantitative evaluations on standard benchmark datasets demonstrate that the proposed method consistently outperforms other state-of-the-art cross-modal retrieval methods.

Fuzzy discretization and control for non-linear, multiple binary input systems

The control of continuous-time linear systems with binary inputs cannot benefit from existing control design techniques because they are based on continuous control actions. In particular, optimal control problems with binary inputs lead to combinatorial optimization problems, which are difficult to solve. In this article we provide an exact discretization model of the binary continuous-time system that results in a non-linear multiple input controlled system. The non-linear model is then converted into a fuzzy discrete Takagi-Sugeno model, thus allowing the use of optimal control techniques based on LMI design. The modelling of the non-linear model by a discrete Takagi-Sugeno model is a complex process but it can be automatically performed as shown in the article and the code of the application examples.