Double-opponent spiking neuron array with orientation selectivity for encoding and spatial-chromatic processing

Science Advances, Год журнала: 2025, Номер 11(7)

Опубликована: Фев. 12, 2025

Color spiking encoding and opponent preprocessing are critical for energy-efficient object perception in the human visual system. Emulating retina brain’s integration of spatial chromatic signals holds promise enhancing efficiency vision sensors. Here, we introduce an artificial neuron array that generates excitatory or inhibitory responses to specific wavelengths with orientation selectivity. The can function as double-opponent receptive fields spatial-chromatic color signals, emulating neural pathway from cortex. With array, recognition accuracy is improved almost twofold compared direct underexposure objects, noise robustness also strengthened. This architecture leverages biological mechanisms simultaneous spike antagonistic information, offering potential highly efficient neuromorphic systems.

Язык: Английский

---

Ultrabroadband Detection and Self-Powered Functionality in Quasi-One-Dimensional Nb₃Se₁₂I Nanowire Photodetectors for Bionic Vision Applications

ACS Applied Materials & Interfaces, Год журнала: 2025, Номер unknown

Опубликована: Март 27, 2025

The burgeoning fields of the Internet things (IoT) and artificial intelligence (AI) have escalated demands for image sensing technologies, necessitating advancements in sensor efficiency functionality. Traditional sensors, structured on von Neumann architectures with discrete processing units, face challenges, such as high power consumption, latency, hardware costs. In this work, we introduced a unique approach through development quasi-one-dimensional nanowire Nb3Se12I-based double-ended photosensor. advanced not only replicated adaptive behavior biological vision systems but also effectively managed decreased sensitivity triggered by intense light stimuli. integration photothermoelectric bolometric effects allows device to operate self-powered mode, offering broadband detectivity ranging from visible (405 nm) midwave infrared (4060 nm). Additionally, structure enables an angle-dependent response polarized polarization ratio 1.83. Our findings suggest that biomimetic based Nb3Se12I could enhance capabilities smart optical sensors machine systems.

Язык: Английский

---

A biomimetic multi-sensory neural system with visual-auditory synergy to integration, recognition, and memory

Chemical Engineering Journal, Год журнала: 2025, Номер unknown, С. 162546 - 162546

Опубликована: Апрель 1, 2025

Язык: Английский

---

Integration of Perovskite/Low‐Dimensional Material Heterostructures for Optoelectronics and Artificial Visual Systems

Advanced Functional Materials, Год журнала: 2025, Номер unknown

Опубликована: Апрель 14, 2025

Abstract Heterojunctions combining halide perovskites with low‐dimensional materials are revolutionizing optoelectronic device design by leveraging complementary properties. Halide perovskites, known for their tunable bandgaps, excellent light‐harvesting, and efficient charge carrier mobility, provide a robust foundation photodetectors (PDs) imaging sensors. Low‐dimensional contribute ultrafast enhanced light‐matter interactions, mechanical flexibility. When integrated into heterostructures, these enable precise control over dynamics, leading to significant improvements in efficiency, stability, response speed. This synergy addresses critical challenges optoelectronics, advancing flexible electronics, wearable sensors, high‐sensitivity systems. Ongoing advancements interface engineering material synthesis continually enhancing the reliability operational efficacy of devices across various environmental conditions. Additionally, heterostructures show substantial promise neuromorphic computing, where properties support energy‐efficient, event‐driven data processing. By mimicking adaptive hierarchical nature biological visual systems, they offer new possibilities real‐time image analysis intelligent decision‐making. review highlights latest developments perovskite‐based heterojunctions transformative role bridging gap between artificial vision, driving technologies such as robotics bio‐inspired

Язык: Английский

---

An artificial visual perception system based on ZnO threshold switching neurons with integrated rate and time-to-first-spike coding

Journal of Materials Chemistry C, Год журнала: 2025, Номер unknown

Опубликована: Янв. 1, 2025

An artificial visual system is developed based on ZnO TS neurons, featuring excellent device performance and stable neuron circuit operation. The utilizes rate-time fusion coding strategies to enable efficient accurate recognition.

Язык: Английский

---