Balanced increases in selectivity and tolerance produce constant sparseness along the ventral visual stream.
Balanced increases in selectivity and tolerance produce constant sparseness along the ventral visual stream.
复制标题
DOI：
10.1523/jneurosci.6125-11.2012
复制
发表时间：
2012-07-25
期刊：
影响因子：
--
通讯作者：
DiCarlo JJ
中科院分区：
文献类型：
--
作者：
Rust NC;DiCarlo JJ
虽然流行的说法表明，沿着腹侧视觉处理流的神经元对特定对象的选择性越来越高，但这似乎与以下颞叶皮质(IT)神经元被广泛调节的事实不符。为了探索这一明显的矛盾，我们比较了恒河猴两个腹侧流阶段(V4和IT)的加工过程。我们证实，与V4神经元相比，IT神经元确实对视觉特征的连接具有更高的选择性，并且这种特征连接选择性的增加伴随着对那些特征的身份保持转换(例如，移位、缩放)的容忍度的增加(不变性)。我们在这里报道，平均而言，V4和IT神经元在自然图像的调谐宽度(稀疏性)上是紧密匹配的，并且平均V4或IT神经元将对所有自然图像的~10%产生强健的放电率反应(超过其峰值观察放电率的50%)。我们还观察到，稀疏性与合取选择性正相关，与V4和IT中的容忍度负相关，与选择性构建和不变性构建计算一致，它们相互抵消以产生稀疏性。我们的结果表明，支持对象识别所需的合取选择性构建和不变性构建计算以平衡的方式实现，以在处理的每个阶段保持稀疏性。
While popular accounts suggest that neurons along the ventral visual processing stream become increasingly selective for particular objects, this appears at odds with the fact that inferior temporal cortical (IT) neurons are broadly tuned. To explore this apparent contradiction, we compared processing in two ventral stream stages (V4 and IT) in the rhesus macaque monkey. We confirmed that IT neurons are indeed more selective for conjunctions of visual features than V4 neurons, and that this increase in feature conjunction selectivity is accompanied by an increase in tolerance (“invariance”) to identity-preserving transformations (e.g. shifting, scaling) of those features. We report here that V4 and IT neurons are, on average, tightly matched in their tuning breadth for natural images (“sparseness”), and that the average V4 or IT neuron will produce a robust firing rate response (over 50% of its peak observed firing rate) to ~10% of all natural images. We also observed that sparseness was positively correlated with conjunction selectivity and negatively correlated with tolerance within both V4 and IT, consistent with selectivity-building and invariance-building computations that offset one another to produce sparseness. Our results imply that the conjunction-selectivity-building and invariance-building computations necessary to support object recognition are implemented in a balanced fashion to maintain sparseness at each stage of processing.