Hybrid Quantum-Classical NLP Classification with Compact Semantic Representations: An Experimental Analysis of Representation Compression
本文研究了通过降维方法将高维语句嵌入转换为紧凑表示，以适应量子机器学习处理限制的问题，使用了PCA、NCA和LDA等技术，并在TREC数据集上验证了方法的有效性。
0 citationsRead paper
Selected work
Representative Papers
FPGA-based Lane Detection System incorporating Temperature and Light Control Units
To address the low detection accuracy and poor environmental adaptability of lane path detection in urban roads and robotic tracks, this paper proposes an FPGA-based real-time lane detection system. The system integrates a temperature- and illumination-adaptive control unit to significantly enhance robustness under varying lighting and thermal conditions. It employs a hardware-optimized Sobel edge detection algorithm, supporting 416×416 input resolution and achieving a single-frame processing latency of only 1.17 ms at a 150 MHz clock frequency. The system outputs key metrics in real time: number of lanes, current lane index, and left/right lane boundary coordinates. Compared with conventional software-based or less-optimized hardware approaches, the proposed design achieves high detection accuracy while substantially reducing computational latency. This yields an efficient, stable, and deployable hardware solution for embedded intelligent vehicle path recognition.
0 citationsRead paper
Recent publications
Latest Papers
Hybrid Quantum-Classical NLP Classification with Compact Semantic Representations: An Experimental Analysis of Representation Compression
本文研究了通过降维方法将高维语句嵌入转换为紧凑表示，以适应量子机器学习处理限制的问题，使用了PCA、NCA和LDA等技术，并在TREC数据集上验证了方法的有效性。
0 citationsRead paper
FPGA-based Lane Detection System incorporating Temperature and Light Control Units
To address the low detection accuracy and poor environmental adaptability of lane path detection in urban roads and robotic tracks, this paper proposes an FPGA-based real-time lane detection system. The system integrates a temperature- and illumination-adaptive control unit to significantly enhance robustness under varying lighting and thermal conditions. It employs a hardware-optimized Sobel edge detection algorithm, supporting 416×416 input resolution and achieving a single-frame processing latency of only 1.17 ms at a 150 MHz clock frequency. The system outputs key metrics in real time: number of lanes, current lane index, and left/right lane boundary coordinates. Compared with conventional software-based or less-optimized hardware approaches, the proposed design achieves high detection accuracy while substantially reducing computational latency. This yields an efficient, stable, and deployable hardware solution for embedded intelligent vehicle path recognition.
0 citationsRead paper