Search

Your search keyword '"He, Ben-guo"' showing total 47 results
Start Over Author "He, Ben-guo"
47 results on '"He, Ben-guo"'

34. Toward Wearable Sensors : Advances, Trends, and Challenges

Author

He, Tongyue, Chen, Junxin, He, Ben-Guo, Wang, Wei, Zhu, Zhi-Liang, Lv, Zhihan, He, Tongyue, Chen, Junxin, He, Ben-Guo, Wang, Wei, Zhu, Zhi-Liang, and Lv, Zhihan

Abstract

Sensors suitable for wearable devices have many special characteristics compared to other sensors, such as stability, sensitivity, sensor volume, biocompatibility, and so on. With the development of wearable technology, amazing wearable sensors have attracted a lot of attention, and some researchers have done a large number of technology explorations and reviews. However, previous surveys generally were concerned with a specified application and comprehensively reviewed the computing techniques for the signals required by this application, as well as how computing can promote data processing. There is a gap in the opposite direction, i.e., the fundamental data source actively stimulates application rather than from the application to the data, and computing promotes the acquisition of data rather than data processing. To fill this gap, starting with different parts of the body as the source of signal, the fundamental data sources that can be obtained and detected are explored by combining the three sensing principles, as well as discussing and analyzing the existing and potential applications of machine learning in simplifying sensor designs and the fabrication of sensors.

Published
2023
Full Text
View/download PDF

35. PGA-Net: Polynomial Global Attention Network With Mean Curvature Loss for Lane Detection

Author

Li, Qiankun, Yu, Xianwang, Chen, Junxin, He, Ben-Guo, Wang, Wei, Rawat, Danda B., and Lyu, Zhihan

Abstract

Lane detection is an important task in the field of automatic driving. Since lane lines usually have complex topologies and exist in various complex scenes (e.g., damaged lanes, severe occlusion, etc.), lane detection remains challenging. In this work, we propose a Polynomial Global Attention Network (PGA-Net) for lane detection, which is an end-to-end model for mining global road information and predicting lanes shape parameter formulas simultaneously. We model lane shape with cubic polynomial function and use the transformer-based DETR model to introduce the context information of lanes and roads to better regress the lane parameters. For polynomial curve modeling, we propose Mean Curvature Loss (MCL) to constrain the curvature of the predicted lanes, thereby enhancing the quality of curve lanes prediction. In addition, we design an improved supervision strategy to eliminate information bias between our parametric prediction methods and the labeling methods of lane datasets. Our method achieves state-of-the-art performance on two popular benchmarks (TuSimple and LLAMAS) and a most challenging benchmark (CULane), while exhibiting accelerated speed (>140fps on 3090 GPU, 28.9% improvement in average) and lightweight model size (<3M, an averaged 83.7% reduction). Our code is available at https://github.com/qklee-lz/PGA-Net.

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results