Spiking neural networks (SNNs) are one of the best practices for efficient event-driven object recognition. To achieve high recognition accuracy, existing methods generally accumulate sufficient binary spike signals over long time steps. As a result, the required power consumption can be comparable even to the traditional artificial neural networks (ANNs). This paper introduces CKADisor, leveraging an ANN teacher to guide the direct training of an SNN student under a shorter time-step setting. To mitigate the representation structure mismatch between ANNs and SNNs, we develop a central-kernel aligned distillation strategy that measures layer-wise feature discrepancies and integrates them into the corresponding optimization objectives. Extensive experiments on three popular event datasets demonstrate that our CKADisor achieves higher accuracy than several state-of-the-art methods under a short time-step setting (T=5) using the same architecture.

Graph Convolutional Networks (GCNs) demonstrate significant potential in recommendation systems but face difficulties with the cold-start problem, especially in integrating new nodes during inference. The typical solution leverages meta-learning for few-shot learning, though it often fails to fully capture collaborative filtering between nodes. In this paper, we revisit the node embedding propagation algorithm in GCNs, emphasizing the importance of collaborative filtering and elucidating the relation between high-order and low-order embeddings. Given the substantial interaction data required for training recommendation models, maintaining a simple model structure remains crucial. To address these challenges, we propose the Linear Graph Network (LGNet), which theoretically compresses multi-layer GCNs into a single layer, enabling the embedding of new nodes during inference. Experimental results on benchmark datasets for link prediction and user cold-start tasks demonstrate that LGNet outperforms existing methods. The code will be available at https://github.com/kunbeibei/LGNet.

Autoregressive object tracking considers bounding box prediction as a sequence generation task, the transformer models the contextual relationships between sequences, this approach transforms the design of complex tracking heads into a generative framework. In visual object tracking, there is rich temporal information in the frames before and after the video, which plays a crucial role in capturing changes in the target's appearance. However, under challenging scenarios such as target similarity confusion or significant appearance variations, existing trackers often fail to accurately capture these changes, leading to apparent deviations in predicted bounding boxes. To alleviate this situation, we introduce a dual temporal dynamic fusion module, designed to enhance cues for target appearance changes and enrich the network's ability to establish associations between the target and the search region. Furthermore, we propose a prior state learning module, which models motion trajectories by integrating historical target states with motion equations. By incorporating historical bounding box sequences, this module infers and generates the bounding box sequence for the current frame, effectively capturing appearance changes through the implicit spatiotemporal information across multiple frames. These enhancements significantly improve the model's tracking accuracy in complex scenarios. Experiments conducted on the LaSOT dataset show the effectiveness of the proposed method, achieving an AUC score of$71.3 \%, P_{\text {Norm }}$of 81.1%, and P of 77.6%. Additionally, ablation experiments further confirm the effectiveness of the proposed modules.

With the popularity of the digital human body, monocular three-dimensional (3D) face reconstruction is widely used in fields such as animation and face recognition. Although current methods trained using single-view image sets perform well in monocular 3D face reconstruction tasks, they tend to rely on the constraints of the a priori model or the appearance conditions of the input images, fundamentally because of the inability to propose an effective method to reduce the effects of two-dimensional (2D) ambiguity. To solve this problem, we developed an unsupervised training framework for monocular face 3D reconstruction using rotational cycle consistency. Specifically, to learn more accurate facial information, we first used an autoencoder to factor the input images and applied these factors to generate normalized frontal views. We then proceeded through a differentiable renderer to use rotational consistency to continuously perceive refinement. Our method provided implicit multi-view consistency constraints on the pose and depth information estimation of the input face, and the performance was accurate and robust in the presence of large variations in expression and pose. In the benchmark tests, our method performed more stably and realistically than other methods that used 3D face reconstruction in monocular 2D images.

In information retrieval-based (IR-based) requirements traceability research, a great deal of researches have focused on establishing trace links between requirements and source code. However, as the description styles of source code and requirements are very different, how to better preprocess the code is crucial for the quality of trace link generation. This paper aims to draw empirical conclusions about code feature extraction, annotation importance assessment, and annotation redundancy removal through comprehensive experiments, which impact the quality of trace links generated by IR-based methods between requirements and source code. The results show that when the average annotaion density is higher than 0.2, feature extraction is recommended. Removing redundancy from code with high annotation redundancy can enhance the quality of trace links. The above experiences can help developers to improve the quality of trace link generation and provide them with advice on writing code.

Context: Requirements trace recovery (RTR) is always time-consuming, tedious, and fallible.There has been a growing interest in applying information retrieval (IR) to automate the process of recover trace links between requirements artifacts and other software artifacts.Objective: In this review, our objective is to identify the state-of-the-art of how IR has been explored to automate RTR and provide an overview of the research at the intersection of these two fields.Method: A systematic mapping study has been conducted, searching the main scientific databases.The search retrieved 1587 citations and 34 articles are retained as primary studies.Results: The results show the most active authors and publication distribution.It presents four kinds of IR models and 21 enhancement strategies applied to perform RTR.Besides, the lists of 37 experimental datasets and 9 measures, commonly used together to evaluate IR-based RTR approaches, are provided.Conclusions: Vector Space Model (VSM) and Latent Semantic Index (LSI) are the most two studied IR models used in RTR.CoEST becomes the most popular, convenient and stable source of datasets.Precision and Recall are the most common measures used to evaluate the performance of IR methods.Overall, IR-based RTR is becoming an increasingly mature cross research field.

Image event recognition is different from object recognition, behaviour recognition and scene recognition. Event is a more advanced concept than object, behaviour and scene. Regarding semantics loss in image event recognition, this paper first proposes a WordNet-based optimization algorithm for concept semantics similarity and describes the semantics relations between different concepts by taking account of such following four impact factors in the WordNet tree as concept semantics distance, concept node depth, concept node density and concept semantics overlap ratio. On that basis, an image event recognition algorithm (CS-IER) based on concept score is proposed, while multi-view learning is applied to fuse concept score and inter-conceptual semantics relations. However, if a higher erroneous concept score is given using CNN, multi-view learning will also augment the concept score approximate to its erroneous concept semantics, thereby leading to the distortion of image representation information. To address this problem, CNN is used to extract channel information to obtain the local features of the image, and it is further fused with the optimized concept score features, so as to form the final image representation information and complete the image event recognition. In experiments, the effectiveness of the proposed algorithm on three datasets is verified.

Vision-based behavior recognition is the analysis and recognition of human behavior in video. It has been widely used in many aspects such as multimedia information retrieval, behavior monitoring, and robot perception. This paper uses the Independent Subspace Analysis (ISA) deep network model feature extraction method, which is based on the ISA model and neural network theory, and combines data preprocessing methods, [Formula: see text]-means clustering methods, and Support Vector Machine (SVM) classifiers to achieve video classification and identification of human behavior. The ISA-based deep network model feature extraction method is an unsupervised learning method that can obtain behavior characteristics with good invariance and characterization capabilities in video human behavior. The experiment was conducted on the basis of the Hollywood2 human behavior data set. This experiment was compared with other commonly used human behavior feature extraction and recognition methods. The experimental results validated the effectiveness and advantages of this method in the classification and recognition of human behavior.

In recent years, how to use renewable energy to reduce the energy cost of internet data center (IDC) has been an urgent problem to be solved. More and more solutions are beginning to consider machine learning, but many of the existing methods need to take advantage of some future information, which is difficult to obtain in the actual operation process. In this paper, we focus on reducing the energy cost of IDC by controlling the energy flow of renewable energy without any future information. we propose an efficient energy dynamic control algorithm based on the theory of reinforcement learning, which approximates the optimal solution by learning the feedback of historical control decisions. For the purpose of avoiding overestimation, improving the convergence ability of the algorithm, we use the double [Formula: see text]-method to further optimize. The extensive experimental results show that our algorithm can on average save the energy cost by 18.3% and reduce the rate of grid intervention by 26.2% compared with other algorithms, and thus has good application prospects.