Deep neural networks rival the representation of primate IT cortex for core visual object recognition.
Deep neural networks rival the representation of primate IT cortex for core visual object recognition.
复制标题
DOI：
10.1371/journal.pcbi.1003963
复制
发表时间：
2014-12
影响因子：
4.3
通讯作者：
DiCarlo JJ
中科院分区：
文献类型：
--
作者：
Cadieu CF;Hong H;Yamins DL;Pinto N;Ardila D;Solomon EA;Majaj NJ;DiCarlo JJ
即使在简短的演示中，以及在对象范例、几何变换和背景变化（也称为核心视觉对象识别）发生变化的情况下，灵长类动物视觉系统也能实现出色的视觉对象识别性能。这种非凡的表现是由下颞叶 (IT) 皮层形成的表征介导的。与此同时，机器学习的最新进展导致使用人工深度神经网络 (DNN) 的对象识别模型性能越来越高。然而，目前尚不清楚 DNN 的表征性能是否可以与大脑相媲美。为了准确地进行这样的比较，一个主要的困难是一个统一的指标，该指标要考虑到实验限制，例如噪声量、神经记录位点的数量和试验次数，以及计算限制，例如解码分类器的复杂性和分类器训练示例的数量。在这项工作中，我们进行了直接比较，纠正了这些实验限制和计算考虑因素。作为我们方法的一部分，我们提出了“核分析”的扩展，该分析将泛化准确性作为表示复杂性的函数进行测量。我们的评估表明，与之前的仿生模型不同，最新的 DNN 在视觉对象识别任务上的表征性能可与 IT 皮层相媲美。此外，我们还表明，在表征性能测量上表现良好的模型在与 IT 的表征相似性测量以及预测个体 IT 多单元响应测量上也表现良好。这些 DNN 是否依赖于类似于灵长类视觉系统的计算机制尚待确定，但与之前所有的仿生模型不同，不能仅仅根据表征性能就排除这种可能性。灵长类动物在确定视觉呈现的物体的类别方面表现出色，即使是在简短的呈现中，以及在物体样本、位置、姿势、比例和背景发生变化的情况下。迄今为止，这种行为是人工计算系统无法比拟的。然而，机器学习领域在产生在对象识别基准上表现出色的人工深度神经网络系统方面取得了长足的进步。在这项研究中，我们测量了数千张图像中颞下皮层 (IT) 神经群体的反应，并将神经特征的性能与最新深度神经网络的特征进行了比较。值得注意的是，我们发现最新的人工深度神经网络的性能与信息皮质的性能相当。深度神经网络和 IT 皮层都创建表征空间，其中具有相同类别对象的图像很接近，而具有不同类别对象的图像则相距很远，即使对象样本、位置、姿势、比例和背景存在很大变化。此外，我们表明这些模型中的顶级特征在预测 IT 神经反应本身方面超过了以前的模型。这一结果表明，最新的深度神经网络可能有助于理解灵长类动物的视觉处理。
The primate visual system achieves remarkable visual object recognition performance even in brief presentations, and under changes to object exemplar, geometric transformations, and background variation (a.k.a. core visual object recognition). This remarkable performance is mediated by the representation formed in inferior temporal (IT) cortex. In parallel, recent advances in machine learning have led to ever higher performing models of object recognition using artificial deep neural networks (DNNs). It remains unclear, however, whether the representational performance of DNNs rivals that of the brain. To accurately produce such a comparison, a major difficulty has been a unifying metric that accounts for experimental limitations, such as the amount of noise, the number of neural recording sites, and the number of trials, and computational limitations, such as the complexity of the decoding classifier and the number of classifier training examples. In this work, we perform a direct comparison that corrects for these experimental limitations and computational considerations. As part of our methodology, we propose an extension of “kernel analysis” that measures the generalization accuracy as a function of representational complexity. Our evaluations show that, unlike previous bio-inspired models, the latest DNNs rival the representational performance of IT cortex on this visual object recognition task. Furthermore, we show that models that perform well on measures of representational performance also perform well on measures of representational similarity to IT, and on measures of predicting individual IT multi-unit responses. Whether these DNNs rely on computational mechanisms similar to the primate visual system is yet to be determined, but, unlike all previous bio-inspired models, that possibility cannot be ruled out merely on representational performance grounds. Primates are remarkable at determining the category of a visually presented object even in brief presentations, and under changes to object exemplar, position, pose, scale, and background. To date, this behavior has been unmatched by artificial computational systems. However, the field of machine learning has made great strides in producing artificial deep neural network systems that perform highly on object recognition benchmarks. In this study, we measured the responses of neural populations in inferior temporal (IT) cortex across thousands of images and compared the performance of neural features to features derived from the latest deep neural networks. Remarkably, we found that the latest artificial deep neural networks achieve performance equal to the performance of IT cortex. Both deep neural networks and IT cortex create representational spaces in which images with objects of the same category are close, and images with objects of different categories are far apart, even in the presence of large variations in object exemplar, position, pose, scale, and background. Furthermore, we show that the top-level features in these models exceed previous models in predicting the IT neural responses themselves. This result indicates that the latest deep neural networks may provide insight into understanding primate visual processing.
登录
查看更多内容
影响因子：
16.2
作者：
Kriegeskorte N;Mur M;Ruff DA;Kiani R;Bodurka J;Esteky H;Tanaka K;Bandettini PA
通讯作者：
Bandettini PA
影响因子：
2.9
作者：
Keerthi, SS;Lin, CJ
通讯作者：
Lin, CJ
影响因子：
1.9
作者：
FUKUSHIMA, K
通讯作者：
FUKUSHIMA, K
影响因子：
2.5
作者：
KOBATAKE, E;TANAKA, K
通讯作者：
TANAKA, K
影响因子：
2.5
作者：
Kobatake, E;Wang, G;Tanaka, K
通讯作者：
Tanaka, K