Search

Spiking neural network (SNN) is a biologically-plausible model and exhibits advantages of high computational capability and low power consumption. While the training of deep SNN is still an open problem, which limits the real-world applications of deep SNN. Here we propose a deep SNN architecture named Spiking SiamFC++ for object tracking with end-to-end direct training. Specifically, the AlexNet network is extended in the time domain to extract the feature, and the surrogate gradient function is adopted to realize direct supervised training of the deep SNN. To examine the performance of the Spiking SiamFC++, several tracking benchmarks including OTB2013, OTB2015, VOT2015, VOT2016, and UAV123 are considered. It is found that, the precision loss is small compared with the original SiamFC++. Compared with the existing SNN-based target tracker, e.g., the SiamSNN, the precision (success) of the proposed Spiking SiamFC++ reaches 0.861 (0.644), which is much higher than that of 0.528 (0.443) achieved by the SiamSNN. To our best knowledge, the performance of the Spiking SiamFC++ outperforms the existing state-of-the-art approaches in SNN-based object tracking, which provides a novel path for SNN application in the field of target tracking. This work may further promote the development of SNN algorithms and neuromorphic chips. [ABSTRACT FROM AUTHOR]

We proposed and experimentally demonstrated a simple and novel photonic spiking neuron based on a distributed feedback (DFB) laser subject to side-mode optical pulse injection (SMOPI). The DFB laser chip is designed and fabricated based on asymmetric equivalent π phase shift (π-EPS) with the reconstruction-equivalent-chirp (REC) technique. Under side-mode continuous-wave (CW) optical injection, excitability pulse was experimentally observed during the dominant mode switching process due to the injection-locked effect. Based on the transition between the excitability regime and the side-mode injection locking effect, the controllable and repeatable neuron-like spiking response can be realized when external stimulus pulses are electro-optically modulated on the CW optical carrier. The experimental results show that the spike threshold, temporal integration, and refractory period, which are important spike processing mechanisms in biological neurons, can all be achieved in the optically-injected DFB laser. The experimental findings are also verified numerically with a rate equation model that considers the SMOPI. To the best of our knowledge, this is the first experimental demonstration of a photonic spiking neuron based on a DFB laser sub ject to SMOPI, which holds promise for realizing large-scale photonic spiking neuron arrays for hardware photonic spiking neural network chips. [ABSTRACT FROM AUTHOR]

In this paper, we propose and demonstrate experimentally and numerically a hierarchical time-delay optical reservoir computing (RC) system based on cascaded vertical-cavity surface-emitting lasers (VCSELs) with feedback and multiple injections. The prediction performance characteristics of the hierarchical time-delay RC system based on cascaded VCSELs under different reservoir layers are compared. Evidently, the prediction performance of the hierarchical time-delay RC system is first improved and saturates as the number of reservoir layers increases. Besides, the impacts of key factors on predicting the hierarchical time-delay RC system are also analyzed in detail experimentally and numerically. This proposed hierarchical time-delay RC system based on VCSELs is useful for the further development of RC systems and may be beneficial to improve the ability of RC systems to solve more complex problems. [ABSTRACT FROM AUTHOR]

This work presents a complete conversion scheme for photonic spiking neural networks (SNNs). We verified that the output of an artificial neural network (ANN) trained with the simulated optical activation function can be directly converted into the spike rate of a photonic spiking neuron model. To reveal the feasibility of hardware implementation, we considered the effects of different bit precisions of data and weight, noise level, and bias current mismatch on the converted results. The proposed scheme was evaluated using the Deterding vowel, IRIS, TIDIGITS, and MNIST datasets for pattern recognition, and achieved mean accuracies of 95.80%, 98.67%, 96.19%, and 92.33%, respectively. The proposed scheme can convert an ANN into a photonic SNN with almost no precision loss, and the performance was comparable to that of an ANN trained with the rectified linear unit function. The proposed scheme can enable the high-performance implementation of photonic SNNs. [ABSTRACT FROM AUTHOR]

Conclusion: In conclusion, a scheme for broadband chaos generation is proposed and demonstrated based on a VCSEL with subharmonic modulation optical injection. A 1/2 sub-harmonic modulation signal is injected into the VCSEL to destabilize the P1 state and generate a complex chaos signal with a bandwidth above 20 GHz. Moreover, such a chaos signal has been used as a physical entropy source to produce the physical number. By adopting a 4-LSBs extraction method, we can realize a 320 Gb/s physical number stream with verified randomness. Our scheme may offer a new alternative solution for ultrafast physical number generation using a single VCSEL. [ABSTRACT FROM AUTHOR]

We experimentally design two photonic spike-timing-dependent plasticity (STDP) schemes based on a single vertical-cavity semiconductor optical amplifier (VCSOA) and demonstrate the photonic implementation of STDP characteristics. In the first scheme, a single-polarized optical pulse train is injected into the VCSOA, in which a pair of optical pulses with a time difference is designed to emulate the pre-synaptic and post-synaptic spikes. In the second scheme, dual-polarized optical pulses emulating the pre-synaptic and post-synaptic spikes are injected into the single VCSOA. Furthermore, the effects of the initial wavelength detuning and the power of the input optical pulse on the STDP curve are analyzed. The proposed photonic STDP schemes based on a single VCSOA need relatively low bias current and power consumption, and thus, are ideal optical synaptic devices forming key components in the construction of the photonic neuromorphic computing system. [ABSTRACT FROM AUTHOR]

A bridge damage detector with preserving integrity based on modified Faster region-based convolutional neural network (R-CNN) is proposed for multiple damage types. The methodologies of dataset collection, damage annotation, and anchors generation are modified. The performance for bridge multiple-damage detectors with ResNet50 or ResNet101 as feature extraction network are compared. The results show that, with the modified Faster R-CNN, the mean average precision reaches 84.56% (76.43%) at the intersection-over-union metrics of 0.5 (0.75). We further demonstrate that the localization offset for Faster R-CNN is lower than that of YOLOv3. The modified bridge damage detector enables better detecting performance, and can preserve the damage integrity. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Acupuncture & Tuina Science is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Integrated circuits (ICs) and optoelectronic chips are the foundation stones of the modern information society. The IC industry has been driven by the so-called “Moore’s law” in the past 60 years, and now has entered the post Moore’s law era. In this paper, we review the recent progress of ICs and optoelectronic chips. The research status, technical challenges and development trend of devices, chips and integrated technologies of typical IC and optoelectronic chips are focused on. The main contents include the development law of IC and optoelectronic chip technology, the IC design and processing technology, emerging memory and chip architecture, brain-like chip structure and its mechanism, heterogeneous integration, quantum chip technology, silicon photonics chip technology, integrated microwave photonic chip, and optoelectronic hybrid integrated chip. [ABSTRACT FROM AUTHOR]

We systematically investigate the spiking dynamics of optical neurons consisting of vertical-cavity surface emitting laser with an embedding saturable absorber (VCESL-SA) and the synchronization properties of coupled networks. The amplitude of spikes and the inter-spike interval of the spiking neurons are mainly considered. We also observe cluster synchronization in regular small networks that can be predicted via the adjacent matrix and further investigate the coherence in large-scale networks (up to a thousand nodes) with 4 typical typologies. Besides, the effect of a biologically observed autapse (which acts as a delayed self-feedback of an optical neuron) on the synchronization of network is taken into account. This work provides basic investigations of the spiking dynamics and synchronization properties in networks of coupled optical neurons in comparison of the fundamental researches based on biological neurons and neuron models, which may be helpful in potential implementations of optical neural networks that mimic the biological brains. [ABSTRACT FROM AUTHOR]

A modified supervised learning rule which is suitable for training photonic spiking neural networks (SNN) is proposed for the first time. The proposed learning rule is independent of the time intervals between actual spike and desired spike or between presynaptic spike and postsynaptic spike. Based on the proposed supervised learning rule, 10 digital images are learned in photonic neural network which consists of 30 presynaptic neurons and 10 postsynaptic neurons. Presynaptic and postsynaptic neurons are photonic neurons based on vertical-cavity surface-emitting lasers with an embedded saturable absorber (VCSEL-SA). The results show that 10 digital images are recognized correctly in photonic SNN after enough training. Additionally, the effects of learning rate, the jitters of learning rate, initial weights distribution of SNN and bias current of postsynaptic neurons (VCSELs-SA) on the recognized error are examined carefully based on the proposed learning rule. To the best of our knowledge, such modified supervised learning rule has not yet been reported, which would contribute to training photonic neural networks, and hence is interesting for neuromorphic photonic systems and pattern recognition. [ABSTRACT FROM AUTHOR]

Chaos synchronization properties in a novel hierarchical tree-type optical network consisting of semiconductor lasers (SLs) are investigated numerically. In such network, each child node is asymmetrically mutually coupled with the parent node. Zero-lag synchronization is found among the nodes belonging to the same layer, while the nodes belonging to different layers are not synchronized, which is denoted as hierarchical chaos synchronization (HCS). The effects of injection strength, asymmetrical injection parameters, bias currents and frequency detuning on the performance of HCS and chaotic complexity are studied in detail. Our results show that high-quality HCS can be obtained under moderate parameters. The stability of the network is also investigated systematically. Further, we explore the robustness of HCS by considering an asymmetrical network. Finally, output bit streams are generated from each SL at a tunable rate of up to about 10 Gbps with verified randomness. These results are enlightening to the multi-user chaotic communication and the synchronization mechanism in biological neuronal networks. [ABSTRACT FROM AUTHOR]

A chaos-based public channel image encryption algorithm among three users is proposed, where the random bits (RBs) generated in a star-type chaotic laser network can be well synchronized and are used as the keys. The proposed algorithm is simple and efficient. Firstly, random bits with verified randomness are generated from the synchronized chaotic semiconductor lasers in a star-type network at a rate of 10Gb/s. Next, lower-triangular error-bits detection is employed to delete the different bits among all the parties over the public channel. Based on the synchronized RBs, the XOR operation is used to diffuse the plain image. Then the hash algorithm is used to get the control parameters matrix from the plain image, and 3D cat map is used to confuse the pixel position through the parameters matrix. Finally, the encrypted image is transmitted in the public channel. The performance tests results, such as key sensitivity, histogram, correlation, differential attack, robustness and entropy analysis, show that the suggested algorithm prevents a powerful computational eavesdropper. Besides, the running speed of this algorithm is linear with the size of plain image. These results open possibilities for multi-user secure communication application. [ABSTRACT FROM AUTHOR]

In this paper, a neuromorphic reservoir computing (RC) system with enhanced memory capacity (MC) based on a vertical-cavity surface-emitting laser (VCSEL) subject to double optical feedbacks (DOF) is proposed and investigated numerically. The aim of this study is to explore the MC of the proposed system. For the purpose of comparison, the MC of the VCSEL-based RC system with single optical feedback (SOF) is also taken into account. It is found that, compared with the VCSEL-based RC system subject to SOF, enhanced MC can be obtained for the VCSEL-based RC system with DOF. Besides, the effects of feedback strength, injected strength, frequency detuning as well as injection current on the MC of the VCSEL-based RC system with DOF are considered. Moreover, the influence of feedback delays is also carefully examined. Thus, such proposed VCSEL-based RC system with DOF provides a prospect for the further development of the neuromorphic photonic system based on RC. [ABSTRACT FROM AUTHOR]

We propose and numerically realize an optical spike-timing dependent plasticity (STDP) scheme by using a single vertical-cavity surface-emitting laser (VCSEL). In the scheme, the VCSEL is subjected to an orthogonally-polarized continuous-wave optical injection (OPCWOI) and dual-polarized pulsed optical injections (DPPOI). Based on the widely used spin-flip model, the response spiking dynamics of VCSEL is numerically studied, and then the optical STDP in a single VCSEL is explored. The roles of bias current, the strength of OPCWOI and DPPOI, and the frequency detuning on the optical STDP curve are numerically analyzed. It is found that, by simultaneously utilizing the response spiking dynamics in two orthogonal polarization modes, an optical STDP could be achieved by using a single VCSEL. Furthermore, the weight update of STDP curve can be calculated in real-time. Additionally, the STDP curves can also be controlled by adjusting some controllable parameters. The real-time optical STDP based on a single VCSEL is numerically realized for the first time, which paves the way towards fully VCSELs-based photonic neuromorphic systems with low power consumption. [ABSTRACT FROM AUTHOR]

A simultaneous unidirectional and bidirectional chaos-based optical communication scheme based on a hybrid coupling semiconductor laser system that consists of one center semiconductor laser (CSL) and multiple side semiconductor lasers (SSLs) is proposed numerically. In this scheme, the SSLs oriented in a linear chain mutually couple with the adjacent SSLs and they are subjected to identical unidirectional injections from the CSL which is a chaotic external cavity semiconductor laser. We theoretically analyze the conditions for different types of chaos synchronization based on the symmetry operation mechanism and injection-locking mechanism, and numerically investigate the influences of operation parameters, parameter mismatch robustness, chaos pass filtering effects and communication performance of the hybrid coupling semiconductor laser system. The simulation results demonstrate that with proper selection of the unidirectional and mutual coupling conditions, the SSLs can synchronize with each other isochronally and simultaneously synchronize to the CSL completely or laggardly, which enables the proposed system to achieve a unidirectional broadcasting communication from the CSL to the SSLs and a bidirectional communication among the SSLs simultaneously. The proposed scheme is beneficial to the implementation of optical chaos communication networks. [ABSTRACT FROM AUTHOR]