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

相关概念视频

Nonconscious Mimicry01:13

Nonconscious Mimicry

5.1K
Nonconscious mimicry occurs when individuals alter their mannerisms to match the behaviors and expressions of those nearby, without intention.
5.1K
Observational Learning01:12

Observational Learning

807
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
807

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

Three-Dimensional Metamaterial for Plasmon-Enhanced Raman Scattering at any Excitation Wavelengths from the Visible to Near-Infrared Range.

Analytical chemistry·2020
Same author

Correction to: Predicting and clustering plant CLE genes with a new method developed specifically for short amino acid sequences.

BMC genomics·2020
Same author

Transferrin receptor (TFRC) is essential for meiotic progression during mouse spermatogenesis.

Zygote (Cambridge, England)·2020
Same author

Effect of aspartic acid on the crystallization kinetics of ACP and dentin remineralization.

Journal of the mechanical behavior of biomedical materials·2020
Same author

Model vs. observation discrepancy in aerosol characteristics during a half-year long campaign in Northeast China: The role of biomass burning.

Environmental pollution (Barking, Essex : 1987)·2020
Same author

Multiple Transformations among Anion-based A<sub>2</sub>L<sub>3</sub> Assemblies: Bicapped Trigonal Antiprism A<sub>8</sub>L<sub>12</sub>, Tetrahedron A<sub>4</sub>L<sub>6</sub>, and Triple Helicate A<sub>2</sub>L<sub>3</sub> (A = Anion).

Journal of the American Chemical Society·2020
Same journal

Relation DETR+: Exploring Explicit Position Relation Prior for Dense Prediction.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

RBF++: Quantifying and Optimizing Reasoning Boundaries across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

CAFE: Cross-View Adaptive Fusion and Cluster Center Enhancement for Robust Multi-View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Learning Shape Anchors for Holistic Indoor Scene Understanding.

IEEE transactions on pattern analysis and machine intelligence·2026
查看所有相关文章

相关实验视频

Updated: Jan 12, 2026

Visualizing Visual Adaptation
04:43

Visualizing Visual Adaptation

Published on: April 24, 2017

9.5K

AdaGen:学习图像合成的自适应政策

Zanlin Ni, Yulin Wang, Yeguo Hua

    IEEE transactions on pattern analysis and machine intelligence
    |October 31, 2025
    PubMed
    概括
    此摘要是机器生成的。

    AdaGen动态调度图像生成参数,改善样本质量并降低计算成本. 这种可学习的框架适应单个样本,在生成任务中表现优于静态方法.

    相关实验视频

    Last Updated: Jan 12, 2026

    Visualizing Visual Adaptation
    04:43

    Visualizing Visual Adaptation

    Published on: April 24, 2017

    9.5K

    科学领域:

    • 计算机视觉 计算机视觉
    • 人工智能的人工智能
    • 机器学习 机器学习

    背景情况:

    • 像MaskGIT,扩散和流动模型这样的生成模型将合成分解为步骤.
    • 代需要配置特定的参数 (例如,面罩比,噪音水平).
    • 目前的方法使用静态,手动设计的时间表,缺乏适应性,需要专家调整.

    研究的目的:

    • 介绍AdaGen,一个通用,可学习和样本适应的框架,用于调度代图像生成.
    • 解决生成模型中静态时间表的局限性.
    • 在图像合成中提高性能,灵活性和效率.

    主要方法:

    • 制定计划作为马尔科夫决策过程 (MDP).
    • 采用轻量级政策网络,通过强化学习进行培训,以调整参数.
    • 为有效的政策培训提出对抗性奖励设计.
    • 整合一个推断时间精炼策略和一个忠诚度-多样性权衡机制.

    主要成果:

    • 在五个基准数据集 (ImageNet,MS-COCO,CC3M,LAION-5B) 和四个生成范式中,AdaGen展示了优越性.
    • 在 DiT-XL 上实现了更好的性能,推断成本降低了大约 3 倍.
    • 将VAR的FID从1.92提升到1.59,使用最小的计算开销.

    结论:

    • 在代生成模型中,AdaGen为适应性调度提供了一种灵活和有效的方法.
    • 对抗性奖励设计对于可靠的质量和多样性至关重要.
    • 在复杂的图像合成任务中,AdaGen显著提高了效率和性能.