Predictive coding of natural images by V1 firing rates and rhythmic synchronization.
Predictive coding of natural images by V1 firing rates and rhythmic synchronization.
复制标题
DOI：
10.1016/j.neuron.2022.01.002
复制
发表时间：
2022-04-06
期刊：
影响因子：
16.2
通讯作者：
Vinck M
中科院分区：
文献类型：
--
作者：
Uran C;Peter A;Lazar A;Barnes W;Klon-Lipok J;Shapcott KA;Roese R;Fries P;Singer W;Vinck M
预测编码是大脑自监督学习的重要候选理论。其中心思想是，感觉反应是由自下而上的输入和上下文预测之间的比较产生的，在这个过程中，速率和同步可能扮演不同的角色。我们记录了清醒的猕猴V1，并开发了一种基于自监督生成神经网络的技术来量化自然图像的刺激可预测性。我们发现，神经元放电率主要调制的高阶图像特征，这与人类感知的相似性判断强烈相关的上下文可预测性。相比之下，V1 gamma（γ）同步随着低水平图像特征的上下文可预测性而单调增加，并且只出现在较大的刺激中。因此，γ同步是由高度可压缩和低维的自然图像引起的。具有低可预测性的自然刺激诱导显著的迟发性β同步，可能反映了皮层反馈。我们的研究结果揭示了自然图像的预测编码的同步和发射率的不同作用。用自监督神经网络量化的自然图像中的可预测性V1发射率随着高级别图像特征的可预测性而降低-而不是低级别图像特征的可预测性γ同步随着低级别图像特征的可预测性而增加使用自监督神经网络来量化自然图像中的刺激可预测性，以研究V1信号的上下文依赖性。发射率随着可预测性而降低，特别是高级别图像特征。相比之下，γ同步随着低水平特征的可预测性而增加，并且出现在低维、强可压缩图像中。
Predictive coding is an important candidate theory of self-supervised learning in the brain. Its central idea is that sensory responses result from comparisons between bottom-up inputs and contextual predictions, a process in which rates and synchronization may play distinct roles. We recorded from awake macaque V1 and developed a technique to quantify stimulus predictability for natural images based on self-supervised, generative neural networks. We find that neuronal firing rates were mainly modulated by the contextual predictability of higher-order image features, which correlated strongly with human perceptual similarity judgments. By contrast, V1 gamma (γ)-synchronization increased monotonically with the contextual predictability of low-level image features and emerged exclusively for larger stimuli. Consequently, γ-synchronization was induced by natural images that are highly compressible and low-dimensional. Natural stimuli with low predictability induced prominent, late-onset beta (β)-synchronization, likely reflecting cortical feedback. Our findings reveal distinct roles of synchronization and firing rates in the predictive coding of natural images. Predictability in natural images quantified with self-supervised neural networks V1 firing rates decrease with predictability of high- not low-level image features γ-synchronization increases with predictability of low-level image features Late-onset β-synchronization for natural scenes with low predictability Uran, Peter et al. use self-supervised neural networks to quantify stimulus predictability in natural images to investigate the context-dependence of V1 signals. Firing rates decrease with the predictability, specifically of high-level image features. By contrast, γ-synchronization increases with the predictability of low-level features and emerges for low-dimensional, strongly compressible images.
登录
查看更多内容
影响因子：
16.2
作者：
Bosman CA;Schoffelen JM;Brunet N;Oostenveld R;Bastos AM;Womelsdorf T;Rubehn B;Stieglitz T;De Weerd P;Fries P
通讯作者：
Fries P
影响因子：
2.5
作者：
Cavanaugh, JR;Bair, W;Movshon, JA
通讯作者：
Movshon, JA
影响因子：
16.2
作者：
Bastos AM;Usrey WM;Adams RA;Mangun GR;Fries P;Friston KJ
通讯作者：
Friston KJ
DOI：
10.1073/pnas.1011284108
发表时间：
2011-07-05
影响因子：
11.1
作者：
Buffalo, Elizabeth A.;Fries, Pascal;Desimone, Robert
通讯作者：
Desimone, Robert
DOI：
10.1073/pnas.2014868117
发表时间：
2020-12-08
影响因子：
11.1
作者：
Bastos AM;Lundqvist M;Waite AS;Kopell N;Miller EK
通讯作者：
Miller EK