Training brain-inspired spiking neural networks (SNNs) with multi-timestep backpropagation imposes substantial memory and computational overhead, hindering their scalability and deployment. To address these challenges, we propose a Neuron-Level Back Propagation (NLBP) method, which eliminates the need for temporal unfolding while maintaining competitive performance. We fit the relationship between a neuron’s average input current and its firing rate using an adaptive sigmoid function. Leveraging this mapping, we derive a neuron-level, single-step backpropagation rule that avoids explicit temporal unfolding. This design significantly reduces computational and memory costs when training deep, large-scale SNNs. Furthermore, NLBP unifies the training of integrate-and-fire (IF) and leaky integrate-and-fire (LIF) neuron models within a framework, and supports both soft and hard reset mechanisms, enhancing generality and practicality. Extensive experiments on pattern classification and object detection demonstrate that NLBP achieves competitive accuracy while reducing memory usage by up to 73.07% and training time by 66.04%.

Motion control is one of the most fundamental issues in robotics, with kinematics and dynamics serving as its core components. While most existing control systems rely on general-purpose processors with large areas and high power consumption. This paper proposes a neuromorphic controller with on-chip learning, satisfying the requirements of high control performance and low cost for robot motion control. The proposed controller consists of an Operational Space Control (OSC) unit and a Spiking Neural Networks (SNNs) processing unit, offering kinematic and dynamic motion control across different (4, 6, 7, and 9) Degrees of Freedom (DoF). Under external disturbances, its control precision and the convergence speed are enhanced by 2.83× and 1.78×, respectively, compared to standard proportional integrated-error derivative (PID) OSC controller. The controller is simulated under 40 nm CMOS technology, occupying a core area of 0.755 mm2and consuming 2.4 mW of power at a frequency of 100 MHz. Compared with other chips used for robot motion control, the proposed controller achieves 2.15× and 65× enhancements in core area and power consumption.

This paper presents an efficient near-eye dynamic vision sensor (DVS)-based processor for real-time eye tracking in augmented reality/virtual reality (AR/VR) devices. The processor takes advantage of the sparse event data with fine time resolution from the DVS, addressing the need for high frame-rate, low-power, and accurate eye tracking on wearable devices with extended battery life. Exploiting the inherent sparsity of event data, we propose an event-density-based region of interest (ROI) determination method that operates directly on event stream, which requires$47\times $fewer operations than the traditional methods, effectively overcoming the latency problem caused by the heavy computational loads. To eliminate the issue of decreasing accuracy at the edges of the field of view (FoV), we customized and fine-tuned a neural network for gaze estimation, ensuring uniformly distributed sub-degree accuracy. An estimator with a streamlined output mapping strategy and an adaptive window-sliding convolution scheme is implemented for gaze estimation acceleration. The processor is designed and fabricated in UMC 40-nm LP technology with a core area of 2.52 mm2 and performs end-to-end eye tracking exclusively with the raw event stream from DVS, achieving an average accuracy of 0.91° within a$96^{\circ } \times 64^{\circ }$FoV. Operating at 200 MHz, it achieves a dynamic frame rate of up to 1.2 kHz and requires only$12.7~\mu $J of energy per gaze estimation. By integrating the DVS, the processor enables real-time, low-power, and accurate eye tracking, enhancing the immersive experience on AR/VR devices and offering intuitive and seamless interactions.

This work proposes Spiking-HDC, a spiking neural network (SNN) processing system with hyperdimensional computing (HDC) and its hardware design for domain transfer scenarios. The input data is firstly fed into a two-layer SNN, serving as a feature extractor. It is followed by a HDC classifier to process feature vectors using hypervectors in binary representation. Such a system leverages HDC’s capability of single-pass learning, which can be adopted to rapidly updating but highly similar tasks by fine-tuning the HDC classifier with limited labeled data. By our experiments, the proposed system demonstrates transfer learning accuracy of 94.76%, 87.12% and 94.37% with few-shot samples on N-MNIST, DVS-Gesture and MNIST datasets, respectively. To apply Spiking-HDC model to extreme edge inference tasks, a dedicated processor is designed and implemented. The simulated results in 40 nm CMOS process illustrate that it has 0.88 mm2core area and 1.8 mW power at 100 MHz frequency. In comparison to similar works, it achieves 3.8×-36× inference energy efficiency enhancement.

In neuromorphic processors, simulating large-scale Spiking Neural Networks (SNNs) for cortical models necessitates a significant increase in communication traffic and memory capacity, due to the lack of exploiting the sparsity of connections. Therefore, this paper proposes CorTile, a scalable neuromorphic processing core designed for cortical simulation. We propose a hybrid-mode router that supports Remote Unicast and Local Broadcast (RULB) routing method, leveraging the high local connectivity and low distal connectivity observed in cortical models. This approach achieves reductions of 36.7% in average router load, 40.7% in peak load, 51.2% in average link traffic, 41.7% in peak traffic, respectively, compared to conventional routing methods. Additionally, the proposed Ternary Content Addressable Memory (TCAM)-based Sparse Connection Memory (TSCM) architecture leads to 87.1% reduction in area and a 62.7% reduction in power consumption. These approaches effectively decrease communication traffic and mitigate the quadratic increase in memory requirements, achieving linear growth instead, thus achieving scalability. The proposed CorTile is simulated using UMC 40-nm CMOS process, occupying an area of 5.15 mm2, supporting a maximum of 8k neurons and 64M synapses. Evaluated using a typical macaque cortex model, it consumes 8.25 mW, with the router operating at 200 MHz and the other modules at 100 MHz. This design achieves an average router load of 12.33 Mpackets/s and peak link traffic of 21.16 MB/s. Thanks to the scalability of the proposed processing core that can be tiled into many-core processors, it paves the way for chiplets and multiple chip integration towards a brain-scale neuromorphic computing system.

With the combination of natural language processing and artificial intelligence techniques, medical extractive question-answering (Q&A) provides valuable insights and assists medical professionals in daily work and scientific research, answering medical questions rapidly and accurately, thus holding significant practical significance. Therefore, research on medical extractive Q&A holds significant practical significance. However, the current state of medical extractive question-answering lacks attention to the interaction and prediction layers in the model structure. To address these issues, this paper proposes the Integrating pre-trained multi-layer structural feature information based Bio-BERT (IPMF-Bio-BERT) approach. This method leverages the rich word vector representations generated by the pre-trained Bio-BERT model, incorporating semantic and syntactic structural information to obtain multi-dimensional and complementary interactive feature information. Additionally, we introduce a flexible guidance network based on interactive information, which combines iterative and pointer network techniques to enhance the predictive performance of the question-answering model. We evaluate our proposed model on the specialized biomedical extractive question-answering BioASQ corpus. Experimental results demonstrate that the IPMF-Bio-BERT training strategy enhances the recognition and predictive capabilities of medical extractive Q&A, we establish new state-of-the-art results by outperforming existing approaches.

The liver is an internal organ that is most commonly affected by cancer metastasis. Hepatocellular carcinoma (HCC) is a major cause of cancer-related deaths worldwide. Identifying the histological category of this disease is a crucial step in determining the most appropriate treatment plan. Currently, the classification of HCC primarily relies on the observation of pathologists through a visual microscope, which is undoubtedly subjective. To address this issue, the present study conducted pathological examination of liver tumors based on few-shot learning. In order to achieve automatic identification of the three categories in HCC tissue, we propose a few-shot learning based HCC histological image classification method that combines contrastive learning (CL) using different backbone architectures with several well-known machine learning methods. We evaluate 12 backbone-classifier combinations, and the best performance is achieved by combining CL using DeiT-TinyR as the backbone architecture with support vector machine (SVM), with an accuracy of 74.42% and an F1 score of 74.58% in the 3-way-10-shot setting. Experimental results demonstrate the effectiveness of this method in NB histological image classification.