Author name cluster

Zhen Cao

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

4 papers

1 author row

AAAI Conference 2026 Conference Paper

Optimization Method for Surrogate Function in Spiking Neural Networks Based on Membrane Potential Distribution

Qi Sun
Zhen Cao
Kaige Geng
Ziyi Zhang
Biao Hou

Spiking Neural Networks (SNNs) offer promising energy efficiency and temporal sparsity for edge intelligence, but their training remains difficult due to gradient mismatch, membrane potential drift, and discretization errors. In this paper, we propose a membrane potential-guided surrogate optimization(MPO) framework that dynamically aligns the surrogate function with the membrane potential distribution to enhance the gradient propagation. Specifically, we introduce a KL-divergence-based regularization to stabilize membrane potential dynamics, and an adaptive width constraint to synchronize the surrogate gradient range with neural activity statistics. Additionally, we design a spike discretization error metric and a correction strategy to mitigate temporal discretization effects. Experiments on CIFAR-10, CIFAR-100, and ImageNet show our method achieves 94.76%, 74.20%, and 65.70% top-1 accuracy respectively, while improving gradient stability and energy efficiency. This work provides a principled optimization scheme for robust and scalable SNN training in practical neuromorphic systems.

PDF Details DOI

EAAI Journal 2026 Journal Article

Spatial dependency learning for image-based anomaly detection in engine combustion

Luyun Miao
Dazhi Zhang
Zhen Cao
Zhichang Guo
Yao Li
Xun Yuan
Jangbo Peng
Chaobo Yang

Traditional scramjet anomaly detection methods are constrained by delayed pressure responses and handcrafted features that depend on expert experience. To address this issue, this paper proposes an intelligent situational awareness algorithm for engine anomaly detection based on chemiluminescence imaging of combustion processes. The model learns the spatial dependencies of local features in stable flame images, using a self-supervised learning framework to characterize the feature distribution of normal image patches and identify anomalies as deviations from this distribution. Experimental results demonstrate that the proposed method achieves 100. 0% accuracy and 100. 0% area under the receiver operating characteristic curve (AUROC) at the image level, while 90. 9% accuracy and 94. 8% AUROC at the pixel level. The algorithm is trained solely on normal images and is capable of simultaneously detecting both abnormal states and abnormal regions.

Details DOI

EAAI Journal 2025 Journal Article

Two-stage vision system: Application of multi-perspective object detection network and character recognition network in industrial product classification

Shengjie Jin
Zhen Cao
Chaogang Yu

In industrial automation, precise identification of incoming goods using vision systems significantly enhances production efficiency and reduces shipment errors. To address issues of size discrepancies in target detection and the need for further category code identification, a two-stage visual classification system has been introduced. The first stage presents a novel You Only Look Once-DualSight Fusion Network (YOLO-DSF) model that adopts multi-perspective backbones: one enhances fine-detail extraction via a lightweight space-to-depth downsampling module (LSPDown), and the other captures global features through a lightweight GhostNet structure. A newly developed FocusFusion module (FFM) merges the outputs from these perspectives, reinforcing small-target detection and maintaining the detection capabilities for multi-scale objects. In the detection head, a low-level feature layer is introduced for improved performance on tiny objects, and a lightweight design is applied to control computational overhead, thereby address staircase convergence issues while preserving efficiency. Compared to the baseline model, the YOLO-DSF achieves a Mean Average Precision at Intersection over Union of 0. 50 (mAP@0. 5) of 98. 6 %, an improvement of 4. 4 %. Moreover, tests on the Northeastern University Surface Defect (NEU-DET) public dataset have also demonstrated the superiority of this model. The second stage involves a U-shaped denoising network (UDNet) that acts as a preprocessor in character recognition, effectively reducing background noise and boosting character visibility, resulting in an accuracy improvement of 1. 5 %. Field tests have demonstrated that the system achieves an object detection accuracy of 96. 3 % and a goods classification accuracy of 98. 9 %, thereby verifying its practicality and value in industrial applications.

Details DOI

AAAI Conference 2023 Conference Paper

KT-Net: Knowledge Transfer for Unpaired 3D Shape Completion

Zhen Cao
Wenxiao Zhang
Xin Wen
Zhen Dong
Yu-Shen Liu
Xiongwu Xiao
Bisheng Yang

Unpaired 3D object completion aims to predict a complete 3D shape from an incomplete input without knowing the correspondence between the complete and incomplete shapes. In this paper, we propose the novel KTNet to solve this task from the new perspective of knowledge transfer. KTNet elaborates a teacher-assistant-student network to establish multiple knowledge transfer processes. Specifically, the teacher network takes complete shape as input and learns the knowledge of complete shape. The student network takes the incomplete one as input and restores the corresponding complete shape. And the assistant modules not only help to transfer the knowledge of complete shape from the teacher to the student, but also judge the learning effect of the student network. As a result, KTNet makes use of a more comprehensive understanding to establish the geometric correspondence between complete and incomplete shapes in a perspective of knowledge transfer, which enables more detailed geometric inference for generating high-quality complete shapes. We conduct comprehensive experiments on several datasets, and the results show that our method outperforms previous methods of unpaired point cloud completion by a large margin. Code is available at https://github.com/a4152684/KT-Net.

PDF Details DOI