Perceptual Learning · Vision Training

用科学的方法
训练你的视力

A psychophysics-based vision training system
built on perceptual learning paradigms.

睿眸(OcuWise)把视觉科学实验室里的经典范式——Gabor 斑块辨别、自适应阶梯法、心理测量函数拟合——搬进手机。 每一次训练都产生可测量的数据,每一条曲线都对应一个可解释的统计学模型。 OcuWise brings classic psychophysics paradigms — Gabor-patch discrimination, adaptive staircases and psychometric-function fitting — to a phone. Every trial yields measurable data; every curve maps to an interpretable statistical model.

GABOR PATCH · f = 3.2 c/deg · σ ≈ 0.42° · θ = 155°
2AFC
强制选择范式
Forced choice
3↓1↑
阶梯收敛规则
Staircase rule
71%
收敛正确率
Criterion
≈0.8
阈值补偿系数
Compensation
20+
最小试次数
Min trials
01 / Science

训练的科学原理 · 不只是"练眼"

The scientific basis — beyond "eye exercise"

视觉能力的提升并不只来自眼球的肌肉,而主要来自大脑视觉皮层对信号的提取效率。睿眸的每一次训练,都是一次标准化的心理物理学测量实验。 Improvement arises chiefly from cortical signal extraction, not ocular muscle. Every session is a standardised psychophysical measurement.

Gabor 斑块刺激

Gabor patch stimulus

Gabor 函数是视觉神经科学中描述初级视皮层(V1)简单细胞感受野的最佳数学模型——高斯窗乘以正弦光栅。用它做刺激,等于直接以神经元"偏好"的语言与其对话。

V1 receptive field

自适应阶梯法

Adaptive staircase

采用 3 下 1 上的加权升降规则,让难度实时跟随受试者表现收敛至约 71% 正确率——这是经验上对二选一强制选择(2AFC)任务最高效的采样点。

3-down / 1-up

心理测量函数拟合

Psychometric function

把每次试次的"刺激强度—正确与否"散点拟合成 Weibull 型心理测量函数,从曲线上取 71% 正确率对应的强度即为阈值估计。报告的是一个统计量,不是感觉。

Weibull fit
02 / Model

结果如何被量化 · 从散点到阈值

From raw trials to a threshold estimate

下图是睿眸收敛后输出的典型心理测量函数:横轴为刺激强度(对比度),纵轴为正确应答概率。蓝色散点是实测数据,绿色曲线是 Weibull 拟合,虚线标出 71% 判据与对应阈值。 A typical fitted psychometric function: contrast (x) vs. proportion correct (y). Dots are measured data, the curve is a Weibull fit, the dashed lines mark the 71% criterion and its threshold.

0.00 0.25 0.50 0.75 1.00 刺激强度 / Stimulus intensity (log contrast) proportion correct criterion 71% threshold θ̂
阈值的含义:受试者能以 71% 正确率分辨的最小对比度。它往下走,意味着在同样条件下看得更清楚——这是一个客观、可重复、可统计检验的量。 Threshold = the contrast at which 71% correct is reached. A lower threshold means better discrimination under identical conditions — an objective, repeatable, statistically testable quantity.
03 / References

方法学文献依据 · 公开可查

Peer-reviewed foundations — publicly verifiable

睿眸所用的刺激范式、收敛规则与拟合方法,均源自下列公开发表的同行评议文献。以下条目按第一作者姓氏排序,期刊、卷期与年份均为真实可查信息。 The paradigms used here originate from the following peer-reviewed, publicly verifiable publications.

Dosher, B. A., & Lu, Z.-L.
Perceptual learning reflects external noise filtering and internal noise reduction through channel reweighting
Proceedings of the National Academy of Sciences, 95(23), 13988–13993 (1998)
Gabor, D.
Theory of communication
Journal of the Institution of Electrical Engineers — Part III: Radio and Communication Engineering, 93(26), 429–457 (1946)
Levi, D. M., & Li, R. W.
Perceptual learning as a potential treatment for amblyopia: A mini-review
Vision Research, 49(21), 2535–2549 (2009)
Marcelja, S.
Mathematical description of the responses of simple cortical cells
Journal of the Optical Society of America, 70(11), 1297–1300 (1980)
Polat, U., Ma-Naim, T., Belkin, M., & Sagi, D.
Improving vision in adult amblyopia by perceptual learning
Proceedings of the National Academy of Sciences, 101(17), 6692–6697 (2004)
Watson, A. B., & Pelli, D. G.
QUEST: A Bayesian adaptive psychometric method
Perception & Psychophysics, 33(2), 113–120 (1983)
Wetherill, G. B., & Levitt, H.
Sequential estimation of points on a psychometric function
British Journal of Mathematical and Statistical Psychology, 18(1), 1–10 (1965)
关于引用范围的说明:上述文献用于说明睿眸所采用的刺激模型(Gabor 函数)、自适应采样(阶梯法 / QUEST)与阈值估计(心理测量函数)的学术来源。引用这些文献不等同于其作者为睿眸背书,也不代表上述研究结论可直接等同于睿眸的临床效果。
These citations document the academic provenance of the models and methods adopted; they do not imply endorsement by the authors, nor do they equate to clinical efficacy of this software.
04 / Research

设计所参考的学术方向 · 而非个人背书

Research traditions this design draws upon

睿眸是一个工程开源项目,而非临床产品。以下列出的是本项目在方法设计上所参考的研究领域与学术传统,用于说明技术路线的来源。这些条目描述的是学科方向,不代表任何在世学者对本产品的推荐或背书。 Listed below are research fields — not individuals — whose traditions inform the design. No living scholar's endorsement is implied.

Field 01

视觉心理物理学

Visual psychophysics

研究物理刺激强度与主观知觉之间的定量映射关系。阈值测量、阶梯法与心理测量函数拟合均属该领域的基础工具。

Field 02

知觉学习

Perceptual learning

研究训练如何持久地改变知觉判断能力,以及这种改变发生在皮层处理的哪一阶段、是否具有朝向与空间频率特异性。

Field 03

计算视觉神经科学

Computational neuroscience of vision

以数学模型刻画感受野与通道特性,Gabor 函数即为其中最有代表性的建模范式。

Field 04

自适应实验设计

Adaptive experimental design

以贝叶斯或阶梯规则动态选择下一次刺激参数,在尽量少的试次内获得尽量精确的估计。

诚信声明 / Integrity statement:本页面不展示任何虚构的"专家推荐"或"名人背书"。所引用的全部文献均为公开发表的真实论文,可在 PubMed、Google Scholar 等平台检索验证;文献引用用于说明方法学来源,不构成对本产品功效的临床证明。
05 / Download

下载安装包 · 选择你习惯的网盘

Download — pick your preferred cloud drive

安装包托管在多个网盘,任选其一即可。若某个链接失效,请稍后再试;下载遇到问题可通过页脚邮箱反馈。 Packages are mirrored across several cloud drives. If a link is down, please try another or check back later.

正在加载下载链接…
Loading download mirrors…
安装说明 / Installation:下载 APK 后,在系统设置中允许「安装未知来源应用」,再点击安装包完成安装。
若链接包含提取码,请先记下提取码再打开链接:提取码已标注在每个按钮下方。