Memristive associative learning has gained significant attention for its ability to mimic fundamental biological learning mechanisms while maintaining system simplicity. In this work, we introduce a high-order memristive associative learning framework with a biologically realistic structure. By utilizing memristors as synaptic modules and their state information to bridge different orders of associative learning, our design effectively establishes associations between multiple stimuli and replicates the transient nature of high-order associative learning. In Pavlov’s classical conditioning experiments, our design achieves a 230% improvement in learning efficiency compared to previous works, with memristor power consumption in the synaptic modules remaining below 11 μW. In large-scale image recognition tasks, we utilize a 20×20 memristor array to represent images, enabling the system to recognize and label test images with semantic information at 100% accuracy. This scalability across different tasks highlights the framework’s potential for a wide range of applications, offering enhanced learning efficiency for current memristor-based neuromorphic systems.

Fundus diseases are prevalent causes of visual impairment and blindness worldwide, particularly in regions with limited access to ophthalmologists for timely diagnosis. Current approaches to fundus disease diagnosis heavily rely on expert-annotated data and AI-assisted image analysis, offering advantages, such as improved accuracy and accessibility. However, the dependency on annotated data poses a significant challenge, especially in regions with limited resources. To address this challenge, we propose a label-free general framework based on self-supervised machine learning. We performed feature distillation on a large number of unlabeled fundus images and employed a linear classifier for the detection of different fundus diseases. In validation experiments on the public and external validation fundus data sets, our model surpassed existing supervised approaches, achieving a remarkable increase in the area under the curve (AUC) of 15.7%, and even outperformed individual human experts. Our approach offers a promising solution to the limitations of current diagnostic methods, enhancing the potential for early and accurate detection of fundus diseases in resource-constrained settings.

In recent years, human activity recognition (HAR) technologies in e-health have triggered broad interest. In literature, mainstream works focus on the body's spatial information (i.e. postures) which lacks the interpretation of key bioinformatics associated with movements, limiting the use in applications requiring comprehensively evaluating motion tasks' correctness. To address the issue, in this article, a Wearables-based Multi-column Neural Network (WMNN) for HAR based on multi-sensor fusion and deep learning is presented. Here, the Tai Chi Eight Methods were utilized as an example as in which both postures and muscle activity strengths are significant. The research work was validated by recruiting 14 subjects in total, and we experimentally show 96.9% and 92.5% accuracy for training and testing, for a total of 144 postures and corresponding muscle activities. The method is then provided with a human-machine interface (HMI), which returns users with motion suggestions (i.e. postures and muscle strength). The report demonstrates that the proposed HAR technique can enhance users' self-training efficiency, potentially promoting the development of the HAR area.

In this paper, a novel nanorobotic strategy for non-destructive and direct electrical characterization of as-grown bundles of single-walled carbon nanotubes (SWCNTs) is presented. For this purpose, test patterns of SWCNT bundles having different diameters are grown on a silicon substrate by chemical vapor deposition. A new design of microstructured four-point-probes is proposed and fabricated allowing for direct contacting of vertically aligned bundles of SWCNTs. A nanorobotic setup is upgraded into a dual endeffector system to achieve good electrical contact between four-point-probe and SWCNT bundle and to perform electrical measurements. First experimental results of non-destructive electrical characterization are presented and discussed.

In this paper a CNNs based circuit for the generation of hyperchaotic signals is proposed. The circuit has been developed for applications in secure communication systems. An Saito oscillator has been designed by using a suitable configuration of a four-cells State-Controlled CNNs. A cryptography system based on the Saito oscillator has been implemented by using inverse system synchronization. The proposed circuit implementation and experimental results are given.

Many works have proved the capability of CNN architecture to reproduce complex phenomena described through partial differential equations (PDEs). This array of simple nonlinear systems called cells, for their analog and spatiotemporal distributed method of processing signals, are considered a powerful tool to generate real-time solutions of nonlinear PDEs. A two-layer 2D CNN is able to solve particular PDEs, so called Reaction-Diffusion Equations. This particular topology of CNN has been called RD-CNN. A silicon implementation of this architecture is presented.

For pt. I, see ISCAS 2000 - IEEE International Symposium on Circuits and Systems, May 28-31, 2000, Geneva, Switzerland. The problem of artificial locomotion is known to represent a difficult task when coping with multiactuated systems controlled by one (central) or more (distributed) digital processors. Nevertheless, the studies carried out from biologists revealed that the underlying mechanism in natural locomotion can be revisited in terms of complex dynamic phenomena such as the generation and propagation of autowaves in active media. The paper describes the work carried out in the laboratories of STMicroelectronics with the aim of generating, "on-the-silicon", the same dynamics found in nature. It will be shown how a 2-layer CNN architecture, so-called Reaction-Diffusion CNN (RD-CNN), can be effective in this task, and how it can be programmed and re-configured to reproduce a variety of complex phenomena.

The methodology proposed in the paper applies artificial intelligence (AI) techniques to the modeling and control of some climate variables within a greenhouse. The nonlinear physical phenomena governing the dynamics of temperature and humidity in such systems are, in fact, difficult to model and control using traditional techniques. The paper proposes a framework for the development of soft computing-based controllers in modern greenhouses.

A growing problem in town areas is noise pollution due to the increasing number of vehicles that daily cross cities. A classical approach to model this kind of system is based on numerical regression, but its performance is not satisfactory due to the nonlinearity of the considered model. A suitable approach can be therefore to determine a fuzzy model of the system. There has been a considerable number of studies on fuzzy identification, where fuzzy implications are used to express rules, in this paper the Tagaki-Sugeno approach has been adopted applying a genetic algorithm during the optimization phase. The obtained models are compared with traditional ones showing the suitability of the proposed method.

A new prediction model is proposed to identify the dynamics corresponding to changing parameters in geophysical structures. The methodology has revealed effective features in the way seismic waveforms are analysed, during continuous monitoring, interpreted and reconstructed by the new neuro-fuzzy-based prediction model. The methodology is based on the concepts of fuzzy logic and ARMA models and hence it has been called FARMA.

In the paper a new structure of a Multi-Layer Perceptron, able to deal with quaternion-valued signals, is proposed. A learning algorithm for the proposed Quaternion MLP (QMLP) is also derived. Such a neural network allows one to interpolate functions of a quaternion variable with a smaller number of connections with respect to the corresponding real valued MLP.>