WCWei Cao

Wei Cao

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Wei Cao has a total of 13 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Improved deep convolutional neural network GoogLeNet for human ECG signal classification

    by Wenchao Ren, Wei Cao, Zeyuan Liu

    Electrocardiogram (ECG) is the most commonly used diagnostic method for heart disease. However, in the process of ECG signal acquisition, it is often interfered by noise, which greatly affects the accuracy of ECG signal classification and diagnosis. To improve the preciseness of ECG classification and diagnosis, we propose a method using continuous wavelet transform (CWT) combined with a deep convolutional network GoogLeNet. We used three sets of ECG data from the PhysioNet database for cardiac arrhythmia (ARR), congestive heart failure (CHF), and normal sinus rhythms (NSR). First, we construct time-frequency representations of ECG signals using CWT, which are known as scale maps. Second, we improve GoogLeNet to better recognize ECG images, which has an accuracy of 96.87%. Finally, this paper visualizes the network filter weights from the first convolutional layer and finds the most relevant channel for the original input, and compares the most relevant channel with the original input to analyze the interpretability of deep neural networks.

  • Open Access

    An Application-oblivious Memory Scheduling System for DNN Accelerators

    by Jiansong Li, Xueying Wang, Xiaobing Chen, Guangli Li, Xiao Dong, Peng Zhao, Xianzhi Yu, Yongxin Yang, Wei Cao, Lei Liu, et al.

    Deep Neural Networks (DNNs) tend to go deeper and wider, which poses a significant challenge to the training of DNNs, due to the limited memory capacity of DNN accelerators. Existing solutions for memory-efficient DNN training are densely coupled with the application features of DNN workloads, e.g., layer structures or computational graphs of DNNs are necessary for these solutions. This would result in weak versatility for DNNs with sophisticated layer structures or complicated computation graphs. These schemes usually need to be re-implemented or re-adapted due to the new layer structures or the unusual operators in the computational graphs introduced by these DNNs. In this article, we review the memory pressure issues of DNN training from the perspective of runtime systems and model the memory access behaviors of DNN workloads. We identify the iterative, regularity , and extremalization properties of memory access patterns for DNN workloads. Based on these observations, we propose AppObMem, an application-oblivious memory scheduling system. AppObMem automatically traces the memory behaviors of DNN workloads and schedules the memory swapping to reduce the memory pressure of the device accelerators without the perception of high-level information of layer structures or computation graphs. Evaluations on a variety of DNN models show that, AppObMem obtains 40–60% memory savings with acceptable performance loss. AppObMem is also competitive with other open sourced SOTA schemes.


  • Compiler-assisted Operator Template Library for DNN Accelerators

    by Jiansong Li, Wei Cao, Xiao Dong, Guangli Li, Xueying Wang, Peng Zhao, Lei Liu, Xiaobing Feng


  • ElasticActor: An Actor System with Automatic Granularity Adjustment

    by Peng Zhao, Lei Liu, Wei Cao, Xiao Dong, Jiansong Li, Xiaobing Feng