Spiking neural networks (SNNs) exhibit superior energy efficiency but suffer from limited performance. In this paper, we consider SNNs as ensembles of temporal subnetworks that share architectures and weights, and highlight a crucial issue that affects their performance: excessive differences in initial states (neuronal membrane potentials) across timesteps lead to unstable subnetwork outputs, resulting in degraded performance. To mitigate this, we promote the consistency of the initial membrane potential distribution and output through membrane potential smoothing and temporally adjacent subnetwork guidance, respectively, to improve overall stability and performance. Moreover, membrane potential smoothing facilitates forward propagation of information and backward propagation of gradients, mitigating the notorious temporal gradient vanishing problem. Our method requires only minimal modification of the spiking neurons without adapting the network structure, making our method generalizable and showing consistent performance gains in 1D speech, 2D object, and 3D point cloud recognition tasks. In particular, on the challenging CIFAR10-DVS dataset, we achieved 83.20\% accuracy with only four timesteps. This provides valuable insights into unleashing the potential of SNNs.

Brain-inspired spiking neural networks (SNNs) promise to be a low-power alternative to computationally intensive artificial neural networks (ANNs), although performance gaps persist. Recent studies have improved the performance of SNNs through knowledge distillation, but rely on large teacher models or introduce additional training overhead. In this paper, we show that SNNs can be naturally deconstructed into multiple submodels for efficient self-distillation. We treat each timestep instance of the SNN as a submodel and evaluate its output confidence, thus efficiently identifying the strong and the weak. Based on this strong and weak relationship, we propose two efficient self-distillation schemes: (1) Strong2Weak: During training, the stronger "teacher" guides the weaker "student", effectively improving overall performance. (2) Weak2Strong: The weak serve as the "teacher", distilling the strong in reverse with underlying dark knowledge, again yielding significant performance gains. For both distillation schemes, we offer flexible implementations such as ensemble, simultaneous, and cascade distillation. Experiments show that our method effectively improves the discriminability and overall performance of the SNN, while its adversarial robustness is also enhanced, benefiting from the stability brought by self-distillation. This ingeniously exploits the temporal properties of SNNs and provides insight into how to efficiently train high-performance SNNs.

In cloud storage, symmetric encryption is a common method to protect the confidentiality of volume data. One critical issue in symmetric encryption is the management of volume symmetric keys such as key generation, update and distribution. Many schemes have adopted hierarchical structures based on key derivation to generate and organize the keys. However, the efficient update of these derived and associated keys and the distribution of multiple derived keys have not been well studied. This paper mainly studies in-situ key update and traffic cost of key distribution. First, we redesign the key node structure of our binary key-derivation tree to provide the basis of the in-situ key update. Then, secure in-situ key update algorithms are proposed, in which forward secrecy and backward secrecy are guaranteed. Finally, we propose a minimal key set generation algorithm, which can effectively reduce the communication cost of key distribution. We also describe the key distribution and derivation process. Security analysis and extensive experimental evaluations show the proposed algorithms are secure, efficient and practical.

The progress of brain cognition and learning mechanisms has provided new inspiration for the next generation of artificial intelligence (AI) and provided the biological basis for the establishment of new models and methods. Brain science can effectively improve the intelligence of existing models and systems. Compared with other reviews, this article provides a comprehensive review of brain-inspired deep learning algorithms for learning, perception, and cognition from microscopic, mesoscopic, macroscopic, and super-macroscopic perspectives. First, this article introduces the brain cognition mechanism. Then, it summarizes the existing studies on brain-inspired learning and modeling from the perspectives of neural structure, cognitive module, learning mechanism, and behavioral characteristics. Next, this article introduces the potential learning directions of brain-inspired learning from four aspects: perception, cognition, understanding, and decision-making. Finally, the top-ten open problems that brain-inspired learning, perception, and cognition currently face are summarized, and the next generation of AI technology has been prospected. This work intends to provide a quick overview of the research on brain-inspired AI algorithms and to motivate future research by illuminating the latest developments in brain science.

Neuromorphic object recognition with spiking neural networks (SNNs) is the cornerstone of low-power neuromorphic computing. However, existing SNNs suffer from significant latency, utilizing 10 to 40 timesteps or more, to recognize neuromorphic objects. At low latencies, the performance of existing SNNs is drastically degraded. In this work, we propose the Shrinking SNN (SSNN) to achieve low-latency neuromorphic object recognition without reducing performance. Concretely, we alleviate the temporal redundancy in SNNs by dividing SNNs into multiple stages with progressively shrinking timesteps, which significantly reduces the inference latency. During timestep shrinkage, the temporal transformer smoothly transforms the temporal scale and preserves the information maximally. Moreover, we add multiple early classifiers to the SNN during training to mitigate the mismatch between the surrogate gradient and the true gradient, as well as the gradient vanishing/exploding, thus eliminating the performance degradation at low latency. Extensive experiments on neuromorphic datasets, CIFAR10-DVS, N-Caltech101, and DVS-Gesture have revealed that SSNN is able to improve the baseline accuracy by 6.55% ~ 21.41%. With only 5 average timesteps and without any data augmentation, SSNN is able to achieve an accuracy of 73.63% on CIFAR10-DVS. This work presents a heterogeneous temporal scale SNN and provides valuable insights into the development of high-performance, low-latency SNNs.

Malicious spreading of hate speech on social media hinders the construction of a harmonious online environment, so efficient automatic hate detection become crucial. However, due to the short text form in which most hate speech exists, the semantic features of these texts are relatively sparse, making it difficult for models to learn enough knowledge for classification. Additionally, an increasing number of hate speech tends to use abbreviated and misspelled hate words to evade detection, making it challenging for many hate speech detection methods to capture hidden hate intents. In this paper, we propose a hate speech detection model based on multi-source feature fusion. The model not only takes into account hate concepts and semantic structure information in the target sentence, but also recognizes abbreviated and misspelled hate words by introducing domain-specific word embedding.Ultimately, it dynamically fuses multisource feature through the attention mechanism to enrich the expressive ability of feature vector and realize the effective detection of complex hate speech. We conduct detailed comparison and ablation experiments, and the results prove the effectiveness of our proposed model.

Unsupervised cross-domain scene segmentation approach adapts the source model to the target domain, which utilizes two-stage strategies to minimize the inter-domain and intra-domain gap. However, the accumulation of errors in the previous stages affects the training of the subsequent stages. In this paper, a framework called statistical and structural domain adaptation (SSDA) is proposed to optimize inter-domain and intra-domain adaptation jointly. Firstly, the statistical inter-domain adaptation (StaIA) is proposed to model dynamic subdomains, which continuously adjust seed samples during the process of domain adaptation to mitigate error accumulation. The dynamic subdomains are modeled by exploring Bayesian uncertainty statistics and global balance statistics, which alleviate the imbalance problem in uncertainty estimation. StaIA encourages the model to transfer comprehensive and genuine knowledge through the seed loss for inter-domain adaptation. Secondly, the structural intra-domain adaptation (StrIA) is proposed to align the intra-domain gap among dynamic subdomains by the structural priors. Specifically, the StrIA models structural priors by truncated conditional random field (TruCRF) loss within the neighborhood, which constrains intra-domain semantic consistency to reduce the intra-domain gap. Experimental results demonstrate the effectiveness of the proposed cross-domain scene segmentation approaches on two commonly-used unsupervised domain adaptation benchmarks. The code is available at https://github.com/ChicalH/SSDA.

Domain generalization (DG) is one of the critical issues for deep learning in unknown domains. How to effectively represent domain-invariant context (DIC) is a difficult problem that DG needs to solve. Transformers have shown the potential to learn generalized features, since the powerful ability to learn global context. In this article, a novel method named patch diversity Transformer (PDTrans) is proposed to improve the DG for scene segmentation by learning global multidomain semantic relations. Specifically, patch photometric perturbation (PPP) is proposed to improve the representation of multidomain in the global context information, which helps the Transformer learn the relationship between multiple domains. Besides, patch statistics perturbation (PSP) is proposed to model the feature statistics of patches under different domain shifts, which enables the model to encode domain-invariant semantic features and improve generalization. PPP and PSP can help to diversify the source domain at the patch level and feature level. PDTrans learns context across diverse patches and takes advantage of self-attention to improve DG. Extensive experiments demonstrate the tremendous performance advantages of the PDTrans over state-of-the-art DG methods.

Semantic segmentation is an important yet unsolved problem in aerial scenes understanding. One of the major challenges is the intense variations of scenes and object scales. In this paper, we propose a novel multi-scale aware-relation network (MANet) to tackle this problem in remote sensing. Inspired by the process of human perception of multi-scale information, we explore discriminative and diverse multi-scale representations. For discriminative multi-scale representations, we propose an inter-class and intra-class region refinement method (IIRR) to reduce feature redundancy caused by fusion. IIRR utilizes the refinement maps with intra- and inter-class scale variation to guide multi-scale fine-grained features. Then, we propose multi-scale collaborative learning (MCL) to enhance the diversity of multi-scale feature representations. The MCL constrains the diversity of multi-scale feature network parameters to obtain diverse information. And the segmentation results are rectified according to the dispersion of the multi-level network predictions. In this way, MANet can learn multi-scale features by collaboratively exploiting the correlation among different scales. Extensive experiments on image and video datasets which have large scale variations have demonstrated the effectiveness of our proposed MANet.

Using the convolutional neural network (CNN) method for image emotion recognition is a research hotspot of deep learning. Previous studies tend to use visual features obtained from a global perspective and ignore the role of local visual features in emotional arousal. Moreover, the CNN shallow feature maps contain image content information; such maps obtained from shallow layers directly to describe low-level visual features may lead to redundancy. In order to enhance image emotion recognition performance, an improved CNN is proposed in this work. Firstly, the saliency detection algorithm is used to locate the emotional region of the image, which is served as the supplementary information to conduct emotion recognition better. Secondly, the Gram matrix transform is performed on the CNN shallow feature maps to decrease the redundancy of image content information. Finally, a new loss function is designed by using hard labels and probability labels of image emotion category to reduce the influence of image emotion subjectivity. Extensive experiments have been conducted on benchmark datasets, including FI (Flickr and Instagram), IAPSsubset, ArtPhoto, and Abstract. The experimental results show that compared with the existing approaches, our method has a good application prospect.

Many efficient approaches are proposed to detect building in remote sensing images. In this paper, in order to learning rich building features better, we propose a full convolutional network with dense connection. There contributions are made: 1) To strengthen feature propagation, an improved dense network is introduced to the full convolution network. 2) We have designed top-down short connections to facilitate the fusion of high and low feature information. 3) In addition, we add the weighted cross entropy edge loss function to make the network pay more attention to building edge in detail. Experiments show that the proposed method achieves excellent performance on the remote sensing image data taken by the QuickBird satellite.

Genetic programming (GP) is an illogical method for automatic programming. It shows creativity in discovering a desired program to solve problem, but in essence bases its searching principle on software testing. This paper is dedicated to establishing a novel GP which combines classical GP and formal approaches like Hoare’s logic, model checking, and automaton, etc. The result indicates these methods can collaborate in the framework pretty well. As has been demonstrated by the experiment, they work in a way that preserves their advantages while each compensates for the deficiencies of the other. So, once an approximate program is obtained, we can say with certainty it is correct with respect to its corresponding pre- and post-conditions.