Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关实验视频

Updated: Sep 11, 2025

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks
11:18

Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

Published on: March 2, 2015

10.4K

设计强大的衍射神经网络,改进横向转移容忍度.

Daniil V Soshnikov, Leonid L Doskolovich, Georgy A Motz

    Journal of the Optical Society of America. A, Optics, image science, and vision
    |August 12, 2025
    PubMed
    概括
    此摘要是机器生成的。

    相关概念视频

    Improving Translational Accuracy02:07

    Improving Translational Accuracy

    11.9K
    Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
    11.9K

    您也可能阅读

    相关文章

    通过共同作者、期刊和引用图与本文相关的文章。

    排序
    Same author

    Surface and Interface Engineering in Integrated Photonic Sensors: Performance Trade-Offs, Stability, and Benchmarking.

    Micromachines·2026
    Same author

    Dispersion Engineering and Sensitivity Enhancement in Photonic Crystal Waveguide Sensors: Current Advances and Emerging Challenges.

    Sensors (Basel, Switzerland)·2026
    Same author

    Design of harmonic diffractive optical elements for generating prescribed intensity distributions.

    Applied optics·2026
    Same author

    Fizeau fringes in wedged resonant gratings.

    Optics express·2026
    Same author

    Optical computation of the divergence of a vector field using a metal-dielectric multilayer.

    Optics express·2025
    Same author

    Supporting quadric method for designing diffractive optical elements in the scalar diffraction theory framework.

    Optics express·2025
    Same journal

    Multi-module collaborative optimization-driven fast speckle correlation imaging in variable environments.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    Same journal

    Secrecy performance analysis of NOMA-UWOC systems over a vertically stratified WGG oceanic turbulence channel.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    Same journal

    Backscattering of plane waves in a composite system containing a rough surface and anisotropic scatterers.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    Same journal

    Aspherical surface construction methods based on extended Jacobi polynomials.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    Same journal

    OCT sidelobe suppression method based on dual-path phase sinusoidal modulation and minimum value fusion.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    Same journal

    Optical design concepts using wavelength-selective diffractive optics to enable miniaturized multimodal endoscopic imaging across separated spectral ranges.

    Journal of the Optical Society of America. A, Optics, image science, and vision·2026
    查看所有相关文章

    我们开发了一种方法来设计可承受定位错误的衍射神经网络 (DNN). 这种方法确保了可靠的图像分类,即使在光学射元件 (DOE) 移位的情况下.

    科学领域:

    • 光学和光子学 在光学和光子学.
    • 机器学习 机器学习
    • 计算机成像成像技术

    背景情况:

    • 衍射神经网络 (DNN) 为图像分类提供了一个有前途的方法.
    • 衍射光学元件 (DOE) 的定位错误可以显著降低DNN的性能.
    • 强大的DNN设计对于实际应用至关重要.

    研究的目的:

    • 提出一种新的方法来设计DNN,这些DNN本质上是坚固的,能够承受构成DOE的横移.
    • 开发一种基于梯度的训练方法,以考虑DNN设计期间的定位错误.
    • 为了证明拟议方法在分类手写数字中的有效性.

    主要方法:

    • 代表分类错误作为一个功能依赖于DOE相位函数和横移向量的函数.
    • 使用这个函数的数学预期作为基于梯度的DNN计算的错误函数.
    • 采用蒙特卡洛方法来计算衍生值,相当于随机DOE横移轮班的训练.

    主要成果:

    • 对于错误函数导数的明确表达式被导出.
    • 根据拟议的方法设计的DNN显示出对DOE横向转移的稳定性.
    • 在可见波长下成功对手写数字进行了分类.

    相关实验视频

    Last Updated: Sep 11, 2025

    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks
    11:18

    Closed-loop Neuro-robotic Experiments to Test Computational Properties of Neuronal Networks

    Published on: March 2, 2015

    10.4K

    结论:

    • 拟议的梯度方法使得DNN的设计能够对DOE横向转移具有稳定性.
    • 设计的DNN即使有显著的横向转移 (最多17个波长),也保持了良好的性能.
    • 这项工作有助于开发可靠和实用的衍射光学计算系统.