Efficient Information Processing Research Articles

Ion Intercalation‐Enabled Reconfigurable Photodetector with Low Programming Voltage and Broadband Response Based on Van Der Waals Tin Disulfide

AbstractArtificial neural networks with integrated sensing and computing capabilities, leveraging reconfigurable optoelectronics, can effectively emulate biological neural networks, thereby enabling rapid and efficient information processing. However, realizing reconfigurable photoresponsivity is often blocked by the requirement for high programming voltages and the limits of the detection spectrum range. This greatly restricts the progress of energy‐efficient and precise neuromorphic vision sensing. Herein, a reconfigurable photodetector with low programming voltage and broadband response is presented via in situ intercalation of Cu+ ions into the van der Waals (vdW) gaps of thermoelectric 2D material SnS2. Interestingly, the vdW gaps provide an ionic transport channel with lower energy barriers compared to oxide‐based memristors, resulting in a low programming voltage (0.5 V). Furthermore, reversible conversion of photo‐detection is achieved from photovoltaic to photo‐thermoelectric (PTE) mode via voltage‐controlled ion distribution, which modulates the phonon scattering rate in the neighboring SnS2 layer. As a result, the response spectrum switches from visible (532 nm) to long‐wave infrared (10 µm) with an on/off ratio as high as 104. Thus, dual‐mode conversion and broadband detection functionality in reconfigurable imaging are realized, suggesting a potential pathway for the development of highly energy‐efficient reconfigurable optoelectronics with a spectrum far beyond human vision.

Metaplasticity and Reservoir Computing in Bio‐Realistic Artificial Synapses with Embedded Localized Au‐Nanoparticle‐Based Memristors

AbstractThis research reports on the control of short‐term and long‐term memory in transition metal oxides embedded with localized gold nanoparticles (Au‐NPs). The HfTiOx/TiSiOx switching layer, after the orderly and uniform insertion of Au‐NPs, demonstrates uniform cycle‐to‐cycle DC switching with an ON–OFF ratio >10. Stable low‐resistance states (LRS) and high‐resistance state (HRS) are maintained up to 104 s with multilevel memory characteristics due to the control of oxygen vacancy concentrations. The localized Au‐NPs enhance the local electric field near the HfTiOx/Au‐NP interface, forming controlled conductive filaments, while the high concentration of oxygen vacancies creates a permanent conduction path inside TiSiOx after the electroforming process. The ITO/HfTiOx/Au‐NP/TiSiOx/TaN memristor exhibits stable, controllable gradual bipolar switching and mimics several biological memory functions, including pulse‐width‐dependent plasticity, spike‐timing‐dependent plasticity, pulse‐frequency‐dependent plasticity, and experience‐dependent plasticity. Additionally, a performance of 50k SET/RESET cycles without any significant degradation is achieved and the facilitation of long‐term potentiation/depression are demonstrated. With the help of controlled oxygen vacancy generation on the surface of Au‐NP inside the HfTiOx/TiSiOx switching layer, the memristor can emulate metaplasticity. Evaluation of a reservoir computing system utilizing the volatile switching of the memristor shows efficient processing of temporal data information which is essential for neuromorphic systems.

Parameters of oculomotor activity and cerebral hemodynamics in students with different types of autonomic reactivity while performing cognitive tasks

BACKGROUND: At the current state of human development, with the high rates of knowledge accumulation and accelerated social and technical evolution, studying the psychophysiological features of cognitive activity within a limited time is essential. AIM: To examine oculomotor activity and cerebral hemodynamics parameters in students with different types of autonomic reactivity while performing cognitive tasks. MATERIALS AND METHODS: The study enrolled 110 students of the Northern (Arctic) Federal University named after M.V. Lomonosov (mean age, 19.0±0.5 years). To identify specific features of cognitive activity in various timing conditions, the type of autonomic nervous system reactivity was determined. Eye movements were tracked when solving cognitive tasks at various timing conditions. A rheoencephalogram was simultaneously recorded to assess cerebral hemodynamics. The efficiency of cognitive activity was assessed in various timing conditions with consideration of the identified individual typological features. RESULTS: The participants demonstrated different efficiency levels of cognitive activity when performing cognitive tasks at various timing conditions. Students with a sympathetic-tone type of autonomic reactivity revealed the most efficient and rapid processing of visual information under a time-constraint setting owing to the presence of a more stable relationship between oculomotor activity and cerebral hemodynamics parameters. Participants with parasympathetic-tone reactivity showed the lowest success rates in cognitive activity and the least stable relationships between oculomotor activity and cerebral hemodynamics. CONCLUSIONS: Young individuals with different types of autonomic reactivity exhibited general and specific alterations in cerebral hemodynamics and oculomotor activity when performing cognitive tasks under a time-constraint setting. They showed different efficiency levels of cognitive activity. Thus, human cognitive abilities under stress depend on the reactivity type of the nervous system. The specific changes in oculomotor reactions and cerebral hemodynamics parameters may be considered success markers of cognitive activity. The most successful cognitive activity under limited time conditions is ensured with a more stable statistical relationship between cerebral hemodynamics and oculomotor activity parameters. This is reflected in a stable factor model of visual cognitive activity, regardless of time limitations.

The impact of emotional valence on the spatial scope of visual selective attention

The brain’s ability to prioritize behaviorally relevant sensory inputs (i.e., targets) while ignoring irrelevant distractors is crucial for efficient information processing. However, the role of emotional valence in modulating selective attention remains underexplored. This study examined how positive and negative emotions alter the spatial scope of visual selective attention using a modified Eriksen Flanker task. Participants viewed an emotional face cue (happy, angry, or neutral) randomly positioned on the screen and then identified the shape of a subsequent neutral target (bowtie or diamond) at the cued location. Adjacent stimuli either matched the target shape (congruent) or differed (incongruent). Results showed that happy faces increased susceptibility to distractors (i.e., a larger incongruency effect), suggesting a broadening of attentional scope, while angry faces reduced susceptibility (i.e., a smaller incongruency effect), indicating a narrowing of focus. Importantly, the magnitude of this emotion-driven attention modulation was negatively correlated with participants’ self-reported levels of psychological distress. Participants with higher stress and depression exhibited weaker attention broadening in response to positive cues. Together, the findings provide behavioral evidence of how emotional valence influences attention scope, offering potential insights into the dynamic interplay between psychological distress, emotional processing, and attention modulation.

Advances in Infrared Detectors for In-Memory Sensing and Computing

In-memory sensing and computing devices integrate the functionalities of sensors, memory, and processors, offering advantages such as low power consumption, high bandwidth, and zero latency, making them particularly suitable for simulating synaptic behavior in biological neural networks. As the pace of digital transformation accelerates, the demand for efficient information processing technologies is increasing, and in-memory sensing and computing devices show great potential in AI, machine learning, and edge computing. In recent years, with the continuous advancement of infrared detector technology, infrared in-memory sensing and computing devices have also seen new opportunities for development. This article reviews the latest research progress in infrared in-memory sensing and computing devices. It first introduces the working principles and performance metrics of in-memory sensing and computing devices, then discusses in detail transistors and memristor-type devices with infrared band response, and finally looks forward to the development prospects of the field. Through innovation in new semiconductor materials and structures, the development trajectory of infrared in-memory sensing and computing devices has been significantly expanded, providing new impetus for the development of a new generation of information technology.

Deep Quasi-Recurrent Self-Attention with Dual Encoder-Decoder in Biomedical CT Image Segmentation.

Developing deep learning models for accurate segmentation of biomedical CT images is challenging due to their complex structures, anatomy variations, noise, and unavailability of sufficient labeled data to train the models. There are many models in the literature, but the researchers are yet to be satisfied with their performance in analyzing biomedical Computed Tomography (CT) images. In this article, we pioneer a deep quasi-recurrent self-attention structure that works with a dual encoder-decoder. The proposed novel deep quasi-recurrent self-attention architecture evokes parameter reuse capability that offers consistency in learning and quick convergence of the model. Furthermore, the quasi-recurrent structure leverages the features acquired from the previous time points and elevates the segmentation quality. The model also efficiently addresses long-range dependencies through a selective focus on contextual information and hierarchical representation. Moreover, the dynamic and adaptive operation, incremental and efficient information processing of the deep quasi-recurrent self-attention structure leads to improved generalization across different scales and levels of abstraction. Along with the model, we innovate a new training strategy that fits with the proposed deep quasi-recurrent self-attention architecture. The model performance is evaluated on various publicly available CT scan datasets and compared with state-of-the-art models. The result shows that the proposed model outperforms them in segmentation quality and training speed. The model can assist physicians in improving the accuracy of medical diagnoses.

Thermally Stable Ag2Se Nanowire Network as an Effective In‐Materio Physical Reservoir Computing Device

AbstractThe artificial intelligence (AI) paradigm shifts from software to implementing general‐purpose or application‐specific hardware systems with lower power requirements. This study explored a material physical reservoir consisting of a material random network, called in‐materio physical reservoir computing (RC) to achieve efficient hardware systems. The device, made up of a random, highly interconnected network of nonlinear Ag2Se nanojunctions as reservoir nodes, demonstrated the requisite characteristics of an in‐materio physical reservoir, including but not limited to nonlinear switching, memory, and higher harmonic generation. The power consumption of the in‐materio physical reservoir is 0.07 nW per nanojunctions, confirming its highly efficient information processing system. As a hardware reservoir, the devices successfully performed waveform generation tasks. Finally, a voice classification by an in‐materio physical reservoir is achieved over 80%, comparable to an RC software simulation. In‐materio physical RC with rich nonlinear dynamics has huge potential for next‐generation hardware‐based AI.

(Invited) IGZO-Based Neuromorphic Devices and Their Application in Spiking Neural Networks

Neuromorphic computing, which mimics the mechanism of the human brain, has attracted significant attention as a next-generation computing technology beyond von Neumann architecture owing to its ability to maximize the efficiency of information processing by connecting numerous synapses and neurons in parallel. Research areas ranging from hardware to networks are actively studied to realize neuromorphic computing. In hardware, research is focused on developing devices that can mimic the operations of synapses and neurons based on various materials, including silicon, two-dimensional material, and metal oxide. In particular, indium-gallium-zinc-oxide (IGZO) is one of the promising materials due to its low leakage current, moderate mobility, and CMOS compatibility. Furthermore, an IGZO-based neuromorphic system provides the advantage of low-temperature processing, enabling the implementation of highly integrated systems through monolithic 3D integration and fabrication on various platforms, such as glass and stretchable substrates. Regarding networks, research on spiking neural networks (SNNs) that mirror the biological mechanism of the human brain is primarily reported. SNN processes the information based on event-based spike propagation by using time-based encoding and integrate-and-fire (I&F) behavior, surpassing the energy efficiency of conventional deep neural networks. Consequently, it is expected that highly efficient information processing in various applications, such as image classification and voice recognition, can be achieved by adopting IGZO neuromorphic hardware and SNN. In this study, we introduced our recent research on the charge trap-based IGZO synaptic transistor and integrate-and-fire (I&F) IGZO neuron circuit, along with SNN simulations for the IGZO neuromorphic system. The charge trap-based IGZO synaptic transistor is a suitable candidate for the artificial synapse due to its high weight linearity, wide dynamic range, and stable read/write operations through a 3-terminal structure. However, the synaptic transistor suffers from the inefficiency of charge trapping/de-trapping owing to the n-type IGZO channel that limits the utilization of holes during the charge transport between the charge trap layer (CTL) and the channel layer. To address this issue, we introduced IGZO as a CTL and optimized the IGZO deposition process to make it degenerate, allowing the electrons in the CTL to exist as free electrons for facilitating electron trapping/de-trapping. Using the synaptic transistor adopting the IGZO CTL, we successfully characterized superior long-term potentiation/depression, which are essential properties for a synapse. For an artificial neuron, a circuit capable of emulating I&F behavior is required. We proposed an I&F neuron circuit composed solely of n-type IGZO TFTs with a membrane capacitor and an inverter chain. Using these IGZO synaptic transistors and neuron circuits, we conducted SNN simulations for the IGZO neuromorphic system and evaluated the performance on three datasets: MNIST, fashion-MNIST, and Cifar-10. Additionally, we suggested a method to optimize the SNN system by adjusting the membrane capacitor of neurons to minimize the accuracy drop caused by information loss during the offline learning of SNN. Figure 1

Physical Reservoir Computing Utilizing Ion‐Gating Transistors Operating in Electric Double Layer and Redox Mechanisms

AbstractThe enormous energy consumption of modern machine learning technologies, such as deep learning and generative artificial intelligence, is one of the most critical concerns of the time. To solve this problem, physical reservoir computing, which uses the non‐linear dynamics exhibited by mechanical systems such as materials and devices as a computational resource for highly efficient information processing, has attracted much attention in recent years. In particular, ion‐gated transistors, a group of devices that control electrical conductivity using electrochemical mechanisms such as electric double layers and redox, show very high computational performance with complex and diverse output properties in contrast to their simple structures, due to the complexity of the physical and chemical processes involved. This research provides an overview of physical reservoir computing using ion‐gating transistors, focusing on the materials used, various computational tasks, and operating mechanisms.

The effects of a dual task on gaze behavior examined during a simulated flight in low-time pilots.

Cognitive load can impair an operator's ability to optimally scan and process relevant information that is critical to the safe and successful operation of an aircraft. Since the cognitive demands experienced by pilots fluctuate throughout a given flight due to changes in task demands that range from high to low cognitive load, it has become increasingly important to objectively track and quantify these changes accordingly. The analysis of eye movements has been shown to be a promising method to understand information acquisition, processing efficiency, and how these aspects of cognition impact pilot performance. Therefore, the aim of the current study was to assess the impact of a dual task paradigm on low-time pilot flight performance and gaze behavior during two phases of flight with varying levels of cognitive load. Twenty-two licensed pilots (<350 h) completed simulated flight circuits alongside an auditory oddball task under visual flight rules conditions. Self-reported situation awareness scores and auditory task performance revealed the dual task was more demanding than the single tasks. Flight performance and gaze behavior indicated that primary task performance and information processing remained unaffected. These results suggest that the recruited pilots attained a level of skill proficiency that enabled the efficient deployment of cognitive resources to successfully complete the flying task under states of increased cognitive load. Combined with previous research findings, the results suggest that the effect of secondary tasks depends on the type of tasks used (i.e., simple/choice response tasks, memory recall, etc.). The utility of using a dual task and gaze behavior to probe flight proficiency and information processing efficiency throughout training are discussed.

Self-Rectifying Memristor-Based Reservoir Computing for Real-Time Intrusion Detection in Cybersecurity.

The increasing sophistication of cybersecurity threats, driven by the proliferation of big data and the Internet of Things (IoT), necessitates the development of advanced real-time intrusion detection systems (IDSs). In this study, we present a novel approach that integrates NiO-doped WO3-x/ZnO bilayer self-rectifying memristors (SRMs) within a reservoir computing (RC) framework for IDS applications. The proposed crossbar array architecture exploits the exceptional dynamic properties of SRMs, achieving a classification accuracy of 93.07% on the CSE-CIC-IDS2018 data set, while demonstrating ultrahigh information-processing efficiency. Our approach not only leverages the tunable characteristics of memristors but also addresses the challenge of sneak path currents in large-scale integration, offering a robust and scalable solution for next-generation IDS. This work exemplifies the power of emerging electronics in enhancing cybersecurity through innovative hardware implementations.

Beyond-local neural information processing in neuronal networks

While there is much knowledge about local neuronal circuitry, considerably less is known about how neuronal input is integrated and combined across neuronal networks to encode higher order brain functions. One challenge lies in the large number of complex neural interactions. Neural networks use oscillating activity for information exchange between distributed nodes. To better understand building principles underlying the observation of synchronized oscillatory activity in a large-scale network, we developed a reductionistic neuronal network model. Fundamental building principles are laterally and temporally interconnected virtual nodes (microcircuits), wherein each node was modeled as a local oscillator. By this building principle, the neuronal network model can integrate information in time and space. The simulation gives rise to a wave interference pattern that spreads over all simulated columns in form of a travelling wave. The model design stabilizes states of efficient information processing across all participating neuronal equivalents. Model-specific oscillatory patterns, generated by complex input stimuli, were similar to electrophysiological high-frequency signals that we could confirm in the primate visual cortex during a visual perception task. Important oscillatory model pre-runners, limitations and strength of our reductionistic model are discussed. Our simple scalable model shows unique integration properties and successfully reproduces a variety of biological phenomena such as harmonics, coherence patterns, frequency-speed relationships, and oscillatory activities. We suggest that our scalable model simulates aspects of a basic building principle underlying oscillatory, large-scale integration of information in small and large brains.

Emerging Spatiotemporal Dynamics in Multiterminal Neuromorphic Nanowire Networks Through Conductance Matrices and Voltage Maps

AbstractSelf‐organizing memristive nanowire (NW) networks are promising candidates for neuromorphic‐type data processing in a physical reservoir computing framework because of their collective emergent behavior, which enables spatiotemporal signal processing. However, understanding emergent dynamics in multiterminal networks remains challenging. Here experimental spatiotemporal characterization of memristive NW networks dynamics in multiterminal configuration is reported, analyzing the activation and relaxation of network's global and local conductance, as well as the inherent spatial nonlinear transformation capabilities. Emergent effects are analyzed i) during activation, by investigating the spatiotemporal dynamics of the electric field distribution across the network through voltage mapping; ii) during relaxation, by monitoring the evolution of the conductance matrix of the multiterminal system. The multiterminal approach also allowed monitoring the spatial distribution of nonlinear activity, demonstrating the impact of different network areas on the system's information processing capabilities. Nonlinear transformation tasks are experimentally performed by driving the network into different conductive states, demonstrating the importance of selecting proper operating conditions for efficient information processing. This work allows a better understanding of the local nonlinear dynamics in NW networks and their impact on the information processing capabilities, providing new insights for a rational design of self‐organizing neuromorphic systems.

Latent encoding of movement in primary visual cortex.

Neurons in the primary visual cortex (V1) are classically thought to encode spatial features of visual stimuli through simple population codes: each neuron exhibits a preferred orientation and preferred spatial frequency that are invariant to other aspects of the visual stimulus. Here, we show that this simple rule does not apply to the representation of major features of stimulus motion, including stimulus direction and temporal frequency (TF). We collected an extensive dataset of cat V1 responses to stimuli covarying in orientation, direction, spatial frequency, and TF to assess the extent of motion selectivity. In over half of V1, we found that the preferred direction changed with stimulus TF, revealing four distinct map motifs embedded within V1's functional architecture. Additionally, preferred TF was mostly uniform across the cortical surface. Despite the lack of spatial modulation for the preferred TF map and the lack of invariance for the preferred direction map, we found using convolutional neural networks that direction, TF and stimulus speed can be accurately decoded from V1 responses at all cortical locations. These findings suggest that subtle modulations of V1 activity may convey fine information about stimulus motion, pointing to a novel primary sensory encoding mechanism despite complex co-variation of responses to multiple attributes across V1 neurons.

Low power tactile sensory neuron using nanoparticle-based strain sensor and memristor

Endowing strain sensors with neuromorphic computing capabilities could permit the efficient processing of tactile information on the edge. The realization of such functionalities from a simple circuit without software processing holds promise for attaining skin-based perception. Here, leveraging the intrinsic neuronal plasticity of memristive neurons, various firing patterns induced by the applied strain were demonstrated. More specifically, tonic, bursting, transition from tonic to bursting, adaptive, and nociceptive activities were captured. The implementation of these patterns permits the facile translation of the analog pressure signals into digital spikes, attaining accurate perception of various tactile characteristics. The tactile sensory neuron consisting of an RC circuit was composed of a SiO2-based conductive bridge memristor exhibiting leaky integrate-and-fire properties and a Pt nanoparticles (NPs)-based strain sensor with a gauge factor of ∼270. A dense layer of Pt NPs was also used as the bottom electrode for the memristive element, yielding the manifestation of a threshold switching mode with a switching voltage of only ∼350 mV and an exceptional switching ratio of 107. Our work provides valuable insights for developing low power neurons with tactile feedback for prosthetics and robotics applications.

Edge Computing for Smart-City Human Habitat: A Pandemic-Resilient, AI-Powered Framework

The COVID-19 pandemic has highlighted the need for a robust medical infrastructure and crisis management strategy as part of smart-city applications, with technology playing a crucial role. The Internet of Things (IoT) has emerged as a promising solution, leveraging sensor arrays, wireless communication networks, and artificial intelligence (AI)-driven decision-making. Advancements in edge computing (EC), deep learning (DL), and deep transfer learning (DTL) have made IoT more effective in healthcare and pandemic-resilient infrastructures. DL architectures are particularly suitable for integration into a pandemic-compliant medical infrastructures when combined with medically oriented IoT setups. The development of an intelligent pandemic-compliant infrastructure requires combining IoT, edge and cloud computing, image processing, and AI tools to monitor adherence to social distancing norms, mask-wearing protocols, and contact tracing. The proliferation of 4G and beyond systems including 5G wireless communication has enabled ultra-wide broadband data-transfer and efficient information processing, with high reliability and low latency, thereby enabling seamless medical support as part of smart-city applications. Such setups are designed to be ever-ready to deal with virus-triggered pandemic-like medical emergencies. This study presents a pandemic-compliant mechanism leveraging IoT optimized for healthcare applications, edge and cloud computing frameworks, and a suite of DL tools. The framework uses a composite attention-driven framework incorporating various DL pre-trained models (DPTMs) for protocol adherence and contact tracing, and can detect certain cyber-attacks when interfaced with public networks. The results confirm the effectiveness of the proposed methodologies.

The impact of bank FinTech on green credit allocation: Evidence from China

This paper investigates the impact of bank FinTech on green credit allocation, using a sample of listed firms in China from 2014 to 2020. We construct a bank FinTech index through textual analysis and find that: (1) Banks with more advanced FinTech significantly allocate more green credit to eco-friendly firms. (2) FinTech primarily empowers banks in green credit allocation by expanding business channels, improving information processing efficiency, and strengthening risk management capabilities. (3) Advanced FinTech banks offer more credit to clean firms in traditional energy sectors and favor firms in high-energy-supply regions over high-energy-consumption ones. (4) There exists a quadratic relationship between bank FinTech and credit allocation to heavily polluting firms, characterized by an initial decrease followed by an increase.

Implementation of controlled unitary gates and its application in a remote-controlled quantum gate.

Recently, remote-controlled quantum information processing has been proposed for its applications in secure quantum processing protocols and distributed quantum networks. For remote-controlled quantum gates, the experimental realization of controlled unitary (CU) gates between any quantum gates is an essential task. Here, we propose and experimentally demonstrate a scheme for implementing CU gates between any pair of unitary gates using the polarization and time-bin degrees of freedom of single photons. Then, we experimentally implement remote-controlled single-qubit unitary gates by controlling either the state preparation or measurement of the control qubit with high process fidelities. We believe the proposed remote-controlled quantum gate model can pave the way for secure and efficient quantum information processing.

Temporal mechanisms underlying visual processing bias in peri-hand space.

The immediate space surrounding the hands has often been termed the peri-hand space (PHS), and is characterized by a smaller reaction time (RT), better detection, and enhanced accuracy for stimuli presented in this space, relative to those stimuli presented beyond this space. Such behavioral changes have been explained in terms of a biased allocation of cognitive resources such as perception, attention, and memory, for the efficient processing of information presented in the PHS. However, in two experiments, the current study shows that these cognitive biases seem to have an underlying temporal basis. The first experiment requires participants to perform a temporal bisection task, whereas the second experiment requires them to perform a verbal estimation task when stimuli are presented either near the hands or relatively far. Results from both experiments give evidence for slowing down of temporal mechanisms in the PHS - reflected in the form of temporal dilation for stimuli presented in the PHS relative to those presented further away. The slowing down of time in the PHS seems crucial in giving sufficient temporal allowance for the allocation of cognitive resources to prioritize the processing of information in the PHS. The findings are in line with the early anticipatory mechanisms associated with the PHS and seem to be driven by the switch/gate mechanism, and not the pacemaker component of the attentional gate model of time perception. Thus, the current study tries to integrate the theories of time perception with the peripersonal space literature.

Inverse stochastic resonance in adaptive small-world neural networks.

Inverse stochastic resonance (ISR) is a counterintuitive phenomenon where noise reduces the oscillation frequency of an oscillator to a minimum occurring at an intermediate noise intensity, and sometimes even to the complete absence of oscillations. In neuroscience, ISR was first experimentally verified with cerebellar Purkinje neurons [Buchin et al., PLOS Comput. Biol. 12, e1005000 (2016)]. These experiments showed that ISR enables a locally optimal information transfer between the input and output spike train of neurons. Subsequent studies have further demonstrated the efficiency of information processing and transfer in neural networks with small-world network topology. We have conducted a numerical investigation into the impact of adaptivity on ISR in a small-world network of noisy FitzHugh-Nagumo (FHN) neurons, operating in a bi-metastable regime consisting of a metastable fixed point and a metastable limit cycle. Our results show that the degree of ISR is highly dependent on the value of the FHN model's timescale separation parameter ε. The network structure undergoes dynamic adaptation via mechanisms of either spike-time-dependent plasticity (STDP) with potentiation-/depression-domination parameter P or homeostatic structural plasticity (HSP) with rewiring frequency F. We demonstrate that both STDP and HSP amplify the effect of ISR when ε lies within the bi-stability region of FHN neurons. Specifically, at larger values of ε within the bi-stability regime, higher rewiring frequencies F are observed to enhance ISR at intermediate (weak) synaptic noise intensities, while values of P consistent with depression-domination (potentiation-domination) consistently enhance (deteriorate) ISR. Moreover, although STDP and HSP control parameters may jointly enhance ISR, P has a greater impact on improving ISR compared to F. Our findings inform future ISR enhancement strategies in noisy artificial neural circuits, aiming to optimize local information transfer between input and output spike trains in neuromorphic systems and prompt venues for experiments in neural networks.