As the emerging trend of the graph-based deep learning, Graph Neural Networks (GNNs) recently attract a significant amount of research attention from various domains. However, existing GNN implementations fail to catch up with the evolving GNN architectures, the ever-increasing graph size, and node-embedding dimensionality, thus, suffering from an unsatisfied performance. To break this hurdle, we propose GN-NAdvisor, an efficient runtime system to systematically accelerate GNN applications on GPUs. First, GNNAdvisor spots the graph-structure information (e.g., graph community) as a new driving force to facilitate GNN acceleration. Besides, GNNAdvisor implements a novel yet highly-efficient groupbased workload management tailored for GNN computation to improve the thread-level performance on GPUs. GNNAdvisor further capitalizes on the GPU memory hierarchy for acceleration by gracefully coordinating the execution of GNNs according to the characteristics of the GPU memory structure. Moreover, GNNAdvisor incorporates a Modeling & Estimating strategy to offer sufficient flexibility for automatic performance tuning across various GNN architectures and input datasets. Extensive experiments show that GNNAdvisor provides average 3.02×, 4.36×, and 52.16× speedup over the state-of-the-art GNN execution frameworks, Deep Graph Library, NeuGraph, and GunRock, respectively.

No abstract

No abstract

No abstract

The study of RMB (renminbi ban currency used in China) serial number recogn and more attention in recent years, for reducin improving financial market stability and soc accuracy of RMB recognition relies heavily o which is a challenging problem due to backgrou uneven illumination. In this paper, we present a extracts the RMB characters directly from scan First, two different techniques, namely skew orientation identification are used to detect t contains RMB serial number. Then the detect binarized by a combined thresholding techniq local contrast average method is introduced to characters from the binarization result. T demonstrate that the proposed binarization me other well-known methods. For character extr an overlap-recall rate of 79.68% and an overlap 98.10% respectively.