Explore the vast range of titles available.

Deep feedforward neural network models of vision dominate in both computational neuroscience and engineering. The primate visual system, by contrast, contains abundant recurrent connections. Recurrent signal flow enables recycling of limited computational resources over time, and so might boost the performance of a physically finite brain or model. Here we show: (1) Recurrent convolutional neural network models outperform feedforward convolutional models matched in their number of parameters in large-scale visual recognition tasks on natural images. (2) Setting a confidence threshold, at which recurrent computations terminate and a decision is made, enables flexible trading of speed for accuracy. At a given confidence threshold, the model expends more time and energy on images that are harder to recognise, without requiring additional parameters for deeper computations. (3) The recurrent model's reaction time for an image predicts the human reaction time for the same image better than several parameter-matched and state-of-the-art feedforward models. (4) Across confidence thresholds, the recurrent model emulates the behaviour of feedforward control models in that it achieves the same accuracy at approximately the same computational cost (mean number of floating-point operations). However, the recurrent model can be run longer (higher confidence threshold) and then outperforms parameter-matched feedforward comparison models. These results suggest that recurrent connectivity, a hallmark of biological visual systems, may be essential for understanding the accuracy, flexibility, and dynamics of human visual recognition.

Deep neural networks (DNNs) excel at visual recognition tasks and are increasingly used as a modeling framework for neural computations in the primate brain. Just like individual brains, each DNN has a unique connectivity and representational profile. Here, we investigate individual differences among DNN instances that arise from varying only the random initialization of the network weights. Using tools typically employed in systems neuroscience, we show that this minimal change in initial conditions prior to training leads to substantial differences in intermediate and higher-level network representations despite similar network-level classification performance. We locate the origins of the effects in an under-constrained alignment of category exemplars, rather than misaligned category centroids. These results call into question the common practice of using single networks to derive insights into neural information processing and rather suggest that computational neuroscientists working with DNNs may need to base their inferences on groups of multiple network instances. Do artificial neural networks, like brains, exhibit individual differences? Using tools from systems neuroscience, this study reveals substantial variability in network-internal representations, calling into question the neuroscientific practice of using single networks as models of brain function.

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 multiregion 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.

Perception is shaped by both incoming sensory input and expectations derived from our prior knowledge. Numerous studies have shown stronger neural activity for surprising inputs, suggestive of predictive processing. However, it is largely unclear what predictions are made across the cortical hierarchy, and therefore what kind of surprise drives this up-regulation of activity. Here, we leveraged fMRI in human volunteers and deep neural network (DNN) models to arbitrate between 2 hypotheses: prediction errors may signal a local mismatch between input and expectation at each level of the cortical hierarchy, or prediction errors may be computed at higher levels and the resulting surprise signal is broadcast to earlier areas in the cortical hierarchy. Our results align with the latter hypothesis. Prediction errors in both low- and high-level visual cortex responded to high-level, but not low-level, visual surprise. This scaling with high-level surprise in early visual cortex strongly diverged from feedforward tuning. Combined, our results suggest that high-level predictions constrain sensory processing in earlier areas, thereby aiding perceptual inference.

Models of fixation selection are a central tool in the quest to understand how the human mind selects relevant information. Using this tool in the evaluation of competing claims often requires comparing different models' relative performance in predicting eye movements. However, studies use a wide variety of performance measures with markedly different properties, which makes a comparison difficult. We make three main contributions to this line of research: First we argue for a set of desirable properties, review commonly used measures, and conclude that no single measure unites all desirable properties. However the area under the ROC curve (a classification measure) and the KL-divergence (a distance measure of probability distributions) combine many desirable properties and allow a meaningful comparison of critical model performance. We give an analytical proof of the linearity of the ROC measure with respect to averaging over subjects and demonstrate an appropriate correction of entropy-based measures like KL-divergence for small sample sizes in the context of eye-tracking data. Second, we provide a lower bound and an upper bound of these measures, based on image-independent properties of fixation data and between subject consistency respectively. Based on these bounds it is possible to give a reference frame to judge the predictive power of a model of fixation selection. We provide open-source python code to compute the reference frame. Third, we show that the upper, between subject consistency bound holds only for models that predict averages of subject populations. Departing from this we show that incorporating subject-specific viewing behavior can generate predictions which surpass that upper bound. Taken together, these findings lay out the required information that allow a well-founded judgment of the quality of any model of fixation selection and should therefore be reported when a new model is introduced.

Our everyday conscious experience of the visual world is fundamentally shaped by the interaction of overt visual attention and object awareness. Although the principal impact of both components is undisputed, it is still unclear how they interact. Here we recorded eye-movements preceding and following conscious object recognition, collected during the free inspection of ambiguous and corresponding unambiguous stimuli. Using this paradigm, we demonstrate that fixations recorded prior to object awareness predict the later recognized object identity, and that subjects accumulate more evidence that is consistent with their later percept than for the alternative. The timing of reached awareness was verified by a reaction-time based correction method and also based on changes in pupil dilation. Control experiments, in which we manipulated the initial locus of visual attention, confirm a causal influence of overt attention on the subsequent result of object perception. The current study thus demonstrates that distinct patterns of overt attentional selection precede object awareness and thereby directly builds on recent electrophysiological findings suggesting two distinct neuronal mechanisms underlying the two phenomena. Our results emphasize the crucial importance of overt visual attention in the formation of our conscious experience of the visual world.

Deep neural networks provide the current best models of visual information processing in the primate brain. Drawing on work from computer vision, the most commonly used networks are pretrained on data from the ImageNet Large Scale Visual Recognition Challenge. This dataset comprises images from 1,000 categories, selected to provide a challenging testbed for automated visual object recognition systems. Moving beyond this common practice, we here introduce ecoset, a collection of >1.5 million images from 565 basic-level categories selected to better capture the distribution of objects relevant to humans. Ecoset categories were chosen to be both frequent in linguistic usage and concrete, thereby mirroring important physical objects in the world. We test the effects of training on this ecologically more valid dataset using multiple instances of two neural network architectures: AlexNet and vNet, a novel architecture designed to mimic the progressive increase in receptive field sizes along the human ventral stream. We show that training on ecoset leads to significant improvements in predicting representations in human higher-level visual cortex and perceptual judgments, surpassing the previous state of the art. Significant and highly consistent benefits are demonstrated for both architectures on two separate functional magnetic resonance imaging (fMRI) datasets and behavioral data, jointly covering responses to 1,292 visual stimuli from a wide variety of object categories. These results suggest that computational visual neuroscience may take better advantage of the deep learning framework by using image sets that reflect the human perceptual and cognitive experience. Ecoset and trained network models are openly available to the research community.

Artificial neural networks (ANNs) inspired by biology are beginning to be widely used to model behavioural and neural data, an approach we call ‘neuroconnectionism’. ANNs have been not only lauded as the current best models of information processing in the brain but also criticized for failing to account for basic cognitive functions. In this Perspective article, we propose that arguing about the successes and failures of a restricted set of current ANNs is the wrong approach to assess the promise of neuroconnectionism for brain science. Instead, we take inspiration from the philosophy of science, and in particular from Lakatos, who showed that the core of a scientific research programme is often not directly falsifiable but should be assessed by its capacity to generate novel insights. Following this view, we present neuroconnectionism as a general research programme centred around ANNs as a computational language for expressing falsifiable theories about brain computation. We describe the core of the programme, the underlying computational framework and its tools for testing specific neuroscientific hypotheses and deriving novel understanding. Taking a longitudinal view, we review past and present neuroconnectionist projects and their responses to challenges and argue that the research programme is highly progressive, generating new and otherwise unreachable insights into the workings of the brain.Artificial neural networks are being widely used to model behavioural and neural data. In this Perspective article, Doerig et al. present neuroconnectionism as a Lakatosian research programme using artificial neural networks as a computational language for expressing falsifiable theories and hypotheses about the brain computations underlying cognition.

Development and aging of the cerebral cortex show similar topographic organization and are governed by the same genes. It is unclear whether the same is true for subcortical regions, which follow fundamentally different ontogenetic and phylogenetic principles. We tested the hypothesis that genetically governed neurodevelopmental processes can be traced throughout life by assessing to which degree brain regions that develop together continue to change together through life. Analyzing over 6000 longitudinal MRIs of the brain, we used graph theory to identify five clusters of coordinated development, indexed as patterns of correlated volumetric change in brain structures. The clusters tended to follow placement along the cranial axis in embryonic brain development, suggesting continuity from prenatal stages, and correlated with cognition. Across independent longitudinal datasets, we demonstrated that developmental clusters were conserved through life. Twin-based genetic correlations revealed distinct sets of genes governing change in each cluster. Single-nucleotide polymorphisms-based analyses of 38,127 cross-sectional MRIs showed a similar pattern of genetic volume–volume correlations. In conclusion, coordination of subcortical change adheres to fundamental principles of lifespan continuity and genetic organization.