The human visual system is an intricate network of brain regions that enables us to recognize the world around us. Despite its abundant lateral and feedback connections, object processing is commonly viewed and studied as a feedforward process. Here, we measure and model the rapid representational dynamics across multiple stages of the human ventral stream using time-resolved brain imaging and deep learning. We observe substantial representational transformations during the first 300 ms of processing within and across ventral-stream regions. Categorical divisions emerge in sequence, cascading forward and in reverse across regions, and Granger causality analysis suggests bidirectional information flow between regions. Finally, recurrent deep neural network models clearly outperform parameter-matched feedforward models in terms of their ability to capture the multi-region cortical dynamics. Targeted virtual cooling experiments on the recurrent deep network models further substantiate the importance of their lateral and top-down connections. These results establish that recurrent models are required to understand information processing in the human ventral stream. deep neural networks, virtual cooling Significance StatementUnderstanding the computational principles that underlie human vision is a key challenge for neuroscience and could help improve machine vision. Feedforward neural network models process their input through a deep cascade of computations. These models can recognize objects in images and explain aspects of human rapid recognition. However, the human brain contains recurrent connections within and between stages of the cascade, which are missing from the models that dominate both engineering and neuroscience. Here we measure and model the dynamics of human brain activity during visual perception. We compare feedforward and recurrent neural network models and find that only recurrent models can account for the dynamic transformations of representations among multiple regions of visual cortex. now published in PNAS: https://www.pnas.org/content/early/

Feedforward neural networks provide the dominant model of how the brain performs visual object recognition. However, these networks lack the lateral and feedback connections, and the resulting recurrent neuronal dynamics, of the ventral visual pathway in the human and non-human primate brain. Here we investigate recurrent convolutional neural networks with bottom-up (B), lateral (L), and top-down (T) connections. Combining these types of connections yields four architectures (B, BT, BL, and BLT), which we systematically test and compare. We hypothesized that recurrent dynamics might improve recognition performance in the challenging scenario of partial occlusion. We introduce two novel occluded object recognition tasks to test the efficacy of the models, digit clutter (where multiple target digits occlude one another) and digit debris (where target digits are occluded by digit fragments). We find that recurrent neural networks outperform feedforward control models (approximately matched in parametric complexity) at recognizing objects, both in the absence of occlusion and in all occlusion conditions. Recurrent networks were also found to be more robust to the inclusion of additive Gaussian noise. Recurrent neural networks are better in two respects: (1) they are more neurobiologically realistic than their feedforward counterparts; (2) they are better in terms of their ability to recognize objects, especially under challenging conditions. This work shows that computer vision can benefit from using recurrent convolutional architectures and suggests that the ubiquitous recurrent connections in biological brains are essential for task performance.