The currently leading artificial neural network models of the visual ventral stream - which are derived from a combination of performance optimization and robustification methods - have demonstrated a remarkable degree of behavioral alignment with humans on visual categorization tasks. We show that image perturbations generated by these models can enhance the ability of humans to accurately report the ground truth class. Furthermore, we find that the same models can also be used out-of-the-box to predict the proportion of correct human responses to individual images, providing a simple, human-aligned estimator of the relative difficulty of each image. Motivated by these observations, we propose to augment visual learning in humans in a way that improves human categorization accuracy at test time. Our learning augmentation approach consists of (i) selecting images based on their model-estimated recognition difficulty, and (ii) applying image perturbations that aid recognition for novice learners. We find that combining these model-based strategies leads to categorization accuracy gains of 33-72% relative to control subjects without these interventions, on unmodified, randomly selected held-out test images. Beyond the accuracy gain, the training time for the augmented learning group was also shortened by 20-23%, despite both groups completing the same number of training trials. We demonstrate the efficacy of our approach in a fine-grained categorization task with natural images, as well as two tasks in clinically relevant image domains - histology and dermoscopy - where visual learning is notoriously challenging. To the best of our knowledge, our work is the first application of artificial neural networks to increase visual learning performance in humans by enhancing category-specific image features.

Studies of the functional role of the primate ventral visual stream have traditionally focused on object categorization, often ignoring -- despite much prior evidence -- its role in estimating "spatial" latents such as object position and pose. Most leading ventral stream models are derived by optimizing networks for object categorization, which seems to imply that the ventral stream is also derived under such an objective. Here, we explore an alternative hypothesis: Might the ventral stream be optimized for estimating spatial latents? And a closely related question: How different -- if at all -- are representations learned from spatial latent estimation compared to categorization? To ask these questions, we leveraged synthetic image datasets generated by a 3D graphic engine and trained convolutional neural networks (CNNs) to estimate different combinations of spatial and category latents. We found that models trained to estimate just a few spatial latents achieve neural alignment scores comparable to those trained on hundreds of categories, and the spatial latent performance of models strongly correlates with their neural alignment. Spatial latent and category-trained models have very similar -- but not identical -- internal representations, especially in their early and middle layers. We provide evidence that this convergence is partly driven by non-target latent variability in the training data, which facilitates the implicit learning of representations of those non-target latents. Taken together, these results suggest that many training objectives, such as spatial latents, can lead to similar models aligned neurally with the ventral stream. Thus, one should not assume that the ventral stream is optimized for object categorization only. As a field, we need to continue to sharpen our measures of comparing models to brains to better understand the functional roles of the ventral stream.

The visual object category reports of artificial neural networks (ANNs) are notoriously sensitive to tiny, adversarial image perturbations. Because human category reports (aka human percepts) are thought to be insensitive to those same small-norm perturbations -- and locally stable in general -- this argues that ANNs are incomplete scientific models of human visual perception. Consistent with this, we show that when small-norm image perturbations are generated by standard ANN models, human object category percepts are indeed highly stable. However, in this very same "human-presumed-stable" regime, we find that robustified ANNs reliably discover low-norm image perturbations that strongly disrupt human percepts. These previously undetectable human perceptual disruptions are massive in amplitude, approaching the same level of sensitivity seen in robustified ANNs. Further, we show that robustified ANNs support precise perceptual state interventions: they guide the construction of low-norm image perturbations that strongly alter human category percepts toward specific prescribed percepts. In sum, these contemporary models of biological visual processing are now accurate enough to guide strong and precise interventions on human perception.

A core problem in visual object learning is using a finite number of images of a new object to accurately identify that object in future, novel images. One longstanding, conceptual hypothesis asserts that this core problem is solved by adult brains through two connected mechanisms: 1) the re-representation of incoming retinal images as points in a fixed, multidimensional neural space, and 2) the optimization of linear decision boundaries in that space, via simple plasticity rules applied to a single downstream layer. Though this scheme is biologically plausible, the extent to which it explains learning behavior in humans has been unclear-in part because of a historical lack of image-computable models of the putative neural space, and in part because of a lack of measurements of human learning behaviors in difficult, naturalistic settings. Here, we addressed these gaps by 1) drawing from contemporary, image-computable models of the primate ventral visual stream to create a large set of testable learning models (n = 2,408 models), and 2) using online psychophysics to measure human learning trajectories over a varied set of tasks involving novel 3D objects (n = 371,000 trials), which we then used to develop (and publicly release) empirical benchmarks for comparing learning models to humans. We evaluated each learning model on these benchmarks, and found those based on deep, high-level representations from neural networks were surprisingly aligned with human behavior. While no tested model explained the entirety of replicable human behavior, these results establish that rudimentary plasticity rules, when combined with appropriate visual representations, have high explanatory power in predicting human behavior with respect to this core object learning problem.

We introduce a visually-guided task-and-motion planning benchmark, which we call the ThreeDWorld Trans-port Challenge. In this challenge, an embodied agent is spawned randomly in a simulated physical home environment and required to transport a small set of objects scattered around the house with containers. We build this benchmark challenge using the ThreeDWorld simulation: a virtual 3D environment where all objects respond to physics, and a robot agent can be controlled using a fully physics-driven navigation and interaction API. We evaluate several existing agents on this benchmark. Experimental results suggest that: 1) a pure RL model struggles on this challenge; 2) state-of-the-art hierarchical planning-based agents can transport some objects but are still far from solving this task. We anticipate that this benchmark will empower researchers to develop more intelligent physics-aware robot learning algorithms.

Humans learn from visual inputs at multiple timescales, both rapidly and flexibly acquiring visual knowledge over short periods, and robustly accumulating online learning progress over longer periods. Modeling these powerful learning capabilities is an important problem for computational visual cognitive science, and models that could replicate them would be of substantial utility in real-world computer vision settings. In this work, we establish benchmarks for both real-time and life-long continual visual learning. Our real-time learning benchmark measures a model's ability to match the rapid visual behavior changes of real humans over the course of minutes and hours, given a stream of visual inputs. Our life-long learning benchmark evaluates the performance of models in a purely online learning curriculum obtained directly from child visual experience over the course of years of development. We evaluate a spectrum of recent deep self-supervised visual learning algorithms on both benchmarks, finding that none of them perfectly match human performance, though some algorithms perform substantially better than others. Interestingly, algorithms embodying recent trends in self-supervised learning -- including BYOL, SwAV and MAE -- are substantially worse on our benchmarks than an earlier generation of self-supervised algorithms such as SimCLR and MoCo-v2. We present analysis indicating that the failure of these newer algorithms is primarily due to their inability to handle the kind of sparse low-diversity datastreams that naturally arise in the real world, and that actively leveraging memory through negative sampling -- a mechanism eschewed by these newer algorithms -- appears useful for facilitating learning in such low-diversity environments. We also illustrate a complementarity between the short and long timescales in the two benchmarks, showing how requiring a single learning algorithm to be locally context-sensitive enough to match real-time learning changes while stable enough to avoid catastrophic forgetting over the long term induces a trade-off that human-like algorithms may have to straddle. Taken together, our benchmarks establish a quantitative way to directly compare learning between neural networks models and human learners, show how choices in the mechanism by which such algorithms handle sample comparison and memory strongly impact their ability to match human learning abilities, and expose an open problem space for identifying more flexible and robust visual self-supervision algorithms.

The computational role of the abundant feedback connections in the ventral visual stream is unclear, enabling humans and nonhuman primates to effortlessly recognize objects across a multitude of viewing conditions. Prior studies have augmented feedforward convolutional neural networks (CNNs) with recurrent connections to study their role in visual processing; however, often these recurrent networks are optimized directly on neural data or the comparative metrics used are undefined for standard feedforward networks that lack these connections. In this work, we develop task-optimized convolutional recurrent (ConvRNN) network models that more correctly mimic the timing and gross neuroanatomy of the ventral pathway. Properly chosen intermediate-depth ConvRNN circuit architectures, which incorporate mechanisms of feedforward bypassing and recurrent gating, can achieve high performance on a core recognition task, comparable to that of much deeper feedforward networks. We then develop methods that allow us to compare both CNNs and ConvRNNs to finely grained measurements of primate categorization behavior and neural response trajectories across thousands of stimuli. We find that high-performing ConvRNNs provide a better match to these data than feedforward networks of any depth, predicting the precise timings at which each stimulus is behaviorally decoded from neural activation patterns. Moreover, these ConvRNN circuits consistently produce quantitatively accurate predictions of neural dynamics from V4 and IT across the entire stimulus presentation. In fact, we find that the highest-performing ConvRNNs, which best match neural and behavioral data, also achieve a strong Pareto trade-off between task performance and overall network size. Taken together, our results suggest the functional purpose of recurrence in the ventral pathway is to fit a high-performing network in cortex, attaining computational power through temporal rather than spatial complexity.

Adversarial examples are often cited by neuroscientists and machine learning researchers as an example of how computational models diverge from biological sensory systems. Recent work has proposed adding biologically-inspired components to visual neural networks as a way to improve their adversarial robustness. One surprisingly effective component for reducing adversarial vulnerability is response stochasticity, like that exhibited by biological neurons. Here, using recently developed geometrical techniques from computational neuroscience, we investigate how adversarial perturbations influence the internal representations of standard, adversarially trained, and biologically-inspired stochastic networks. We find distinct geometric signatures for each type of network, revealing different mechanisms for achieving robust representations. Next, we generalize these results to the auditory domain, showing that neural stochasticity also makes auditory models more robust to adversarial perturbations. Geometric analysis of the stochastic networks reveals overlap between representations of clean and adversarially perturbed stimuli, and quantitatively demonstrate that competing geometric effects of stochasticity mediate a tradeoff between adversarial and clean performance. Our results shed light on the strategies of robust perception utilized by adversarially trained and stochastic networks, and help explain how stochasticity may be beneficial to machine and biological computation.

Current state-of-the-art object recognition models are largely based on convolutional neural network (CNN) architectures, which are loosely inspired by the primate visual system. However, these CNNs can be fooled by imperceptibly small, explicitly crafted perturbations, and struggle to recognize objects in corrupted images that are easily recognized by humans. Here, by making comparisons with primate neural data, we first observed that CNN models with a neural hidden layer that better matches primate primary visual cortex (V1) are also more robust to adversarial attacks. Inspired by this observation, we developed VOneNets, a new class of hybrid CNN vision models. Each VOneNet contains a fixed weight neural network front-end that simulates primate V1, called the VOneBlock, followed by a neural network back-end adapted from current CNN vision models. The VOneBlock is based on a classical neuroscientific model of V1: the linear-nonlinear-Poisson model, consisting of a biologically-constrained Gabor filter bank, simple and complex cell nonlinearities, and a V1 neuronal stochasticity generator. After training, VOneNets retain high ImageNet performance, but each is substantially more robust, outperforming the base CNNs and state-of-the-art methods by 18% and 3%, respectively, on a conglomerate benchmark of perturbations comprised of white box adversarial attacks and common image corruptions. Finally, we show that all components of the VOneBlock work in synergy to improve robustness. While current CNN architectures are arguably brain-inspired, the results presented here demonstrate that more precisely mimicking just one stage of the primate visual system leads to new gains in ImageNet-level computer vision applications.

Much of the recent improvement in neural networks for computer vision has resulted from discovery of new networks architectures. Most prior work has used the performance of candidate models following limited training to automatically guide the search in a feasible way. Could further gains in computational efficiency be achieved by guiding the search via measurements of a high performing network with unknown detailed architecture (e.g. the primate visual system)? As one step toward this goal, we use representational similarity analysis to evaluate the similarity of internal activations of candidate networks with those of a (fixed, high performing) teacher network. We show that adopting this evaluation metric could produce up to an order of magnitude in search efficiency over performance-guided methods. Our approach finds a convolutional cell structure with similar performance as was previously found using other methods but at a total computational cost that is two orders of magnitude lower than Neural Architecture Search (NAS) and more than four times lower than progressive neural architecture search (PNAS). We further show that measurements from only ~300 neurons from primate visual system provides enough signal to find a network with an Imagenet top-1 error that is significantly lower than that achieved by performance-guided architecture search alone. These results suggest that representational matching can be used to accelerate network architecture search in cases where one has access to some or all of the internal representations of a teacher network of interest, such as the brain's sensory processing networks.

Deep convolutional artificial neural networks (ANNs) are the leading class of candidate models of the mechanisms of visual processing in the primate ventral stream. While initially inspired by brain anatomy, over the past years, these ANNs have evolved from a simple eight-layer architecture in AlexNet to extremely deep and branching architectures, demonstrating increasingly better object categorization performance, yet bringing into question how brain-like they still are. In particular, typical deep models from the machine learning community are often hard to map onto the brain's anatomy due to their vast number of layers and missing biologically-important connections, such as recurrence. Here we demonstrate that better anatomical alignment to the brain and high performance on machine learning as well as neuroscience measures do not have to be in contradiction. We developed CORnet-S, a shallow ANN with four anatomically mapped areas and recurrent connectivity, guided by Brain-Score, a new large-scale composite of neural and behavioral benchmarks for quantifying the functional fidelity of models of the primate ventral visual stream. Despite being significantly shallower than most models, CORnet-S is the top model on Brain-Score and outperforms similarly compact models on ImageNet. Moreover, our extensive analyses of CORnet-S circuitry variants reveal that recurrence is the main predictive factor of both Brain-Score and ImageNet top-1 performance. Finally, we report that the temporal evolution of the CORnet-S "IT" neural population resembles the actual monkey IT population dynamics. Taken together, these results establish CORnet-S, a compact, recurrent ANN, as the current best model of the primate ventral visual stream.

Feed-forward convolutional neural networks (CNNs) are currently state-of-the-art for object classification tasks such as ImageNet. Further, they are quantitatively accurate models of temporally-averaged responses of neurons in the primate brain's visual system. However, biological visual systems have two ubiquitous architectural features not shared with typical CNNs: local recurrence within cortical areas, and long-range feedback from downstream areas to upstream areas. Here we explored the role of recurrence in improving classification performance. We found that standard forms of recurrence (vanilla RNNs and LSTMs) do not perform well within deep CNNs on the ImageNet task. In contrast, novel cells that incorporated two structural features, bypassing and gating, were able to boost task accuracy substantially. We extended these design principles in an automated search over thousands of model architectures, which identified novel local recurrent cells and long-range feedback connections useful for object recognition. Moreover, these task-optimized ConvRNNs matched the dynamics of neural activity in the primate visual system better than feedforward networks, suggesting a role for the brain's recurrent connections in performing difficult visual behaviors.

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.

Humans recognize visually-presented objects rapidly and accurately. To understand this ability, we seek to construct models of the ventral stream, the series of cortical areas thought to subserve object recognition. One tool to assess the quality of a model of the ventral stream is the Representation Dissimilarity Matrix (RDM), which uses a set of visual stimuli and measures the distances produced in either the brain (i.e. fMRI voxel responses, neural firing rates) or in models (features). Previous work has shown that all known models of the ventral stream fail to capture the RDM pattern observed in either IT cortex, the highest ventral area, or in the human ventral stream. In this work, we construct models of the ventral stream using a novel optimization procedure for category-level object recognition problems, and produce RDMs resembling both macaque IT and human ventral stream. The model, while novel in the optimization procedure, further develops a long-standing functional hypothesis that the ventral visual stream is a hierarchically arranged series of processing stages optimized for visual object recognition.

The anterior inferotemporal cortex (IT) is the highest stage along the hierarchy of visual areas that, in primates, processes visual objects. Although several lines of evidence suggest that IT primarily represents visual shape information, some recent studies have argued that neuronal ensembles in IT code the semantic membership of visual objects (i.e., represent conceptual classes such as animate and inanimate objects). In this study, we investigated to what extent semantic, rather than purely visual information, is represented in IT by performing a multivariate analysis of IT responses to a set of visual objects. By relying on a variety of machine-learning approaches (including a cutting-edge clustering algorithm that has been recently developed in the domain of statistical physics), we found that, in most instances, IT representation of visual objects is accounted for by their similarity at the level of shape or, more surprisingly, low-level visual properties. Only in a few cases we observed IT representations of semantic classes that were not explainable by the visual similarity of their members. Overall, these findings reassert the primary function of IT as a conveyor of explicit visual shape information, and reveal that low-level visual properties are represented in IT to a greater extent than previously appreciated. In addition, our work demonstrates how combining a variety of state-of-the-art multivariate approaches, and carefully estimating the contribution of shape similarity to the representation of object categories, can substantially advance our understanding of neuronal coding of visual objects in cortex.

Tolerance (“invariance”) to identity-preserving image variation (e.g. variation in position, scale, pose, illumination) is a fundamental problem that any visual object recognition system, biological or engineered, must solve. While standard natural image database benchmarks are useful for guiding progress in computer vision, they can fail to probe the ability of a recognition system to solve the invariance problem. Thus, to understand which computational approaches are making progress on solving the invariance problem, we compared and contrasted a variety of state-of-the-art visual representations using synthetic recognition tasks designed to systematically probe invariance. We successfully re-implemented a variety of state-of-the-art visual representations and confirmed their published performance on a natural image benchmark. We here report that most of these representations perform poorly on invariant recognition, but that one representation shows significant performance gains over two baseline representations. We also show how this approach can more deeply illuminate the strengths and weaknesses of different visual representations and thus guide progress on invariant object recognition.

In recent years, large databases of natural images have become increasingly popular in the evaluation of face and object recognition algorithms. However, Pinto et al. previously illustrated an inherent danger in using such sets, showing that an extremely basic recognition system, built on a trivial feature set, was able to take advantage of low-level regularities in popular object and face recognition sets, performing on par with many state-of-the-art systems. Recently, several groups have raised the performance “bar” for these sets, using more advanced classification tools. However, it is difficult to know whether these improvements are due to progress towards solving the core computational problem, or are due to further improvements in the exploitation of low-level regularities. Here, we show that even modest optimization of the simple model introduced by Pinto et al. using modern multiple kernel learning (MKL) techniques once again yields “state-of-the-art” performance levels on a standard face recognition set (“labeled faces in the wild”). However, at the same time, even with the inclusion of MKL techniques, systems based on these simple features still fail on a synthetic face recognition test that includes more “realistic” view variation by design. These results underscore the importance of building test sets focussed on capturing the central computational challenges of real-world face recognition.

While many models of biological object recognition share a common set of "broad-stroke" properties, the performance of any one model depends strongly on the choice of parameters in a particular instantiation of that model--e.g., the number of units per layer, the size of pooling kernels, exponents in normalization operations, etc. Since the number of such parameters (explicit or implicit) is typically large and the computational cost of evaluating one particular parameter set is high, the space of possible model instantiations goes largely unexplored. Thus, when a model fails to approach the abilities of biological visual systems, we are left uncertain whether this failure is because we are missing a fundamental idea or because the correct "parts" have not been tuned correctly, assembled at sufficient scale, or provided with enough training. Here, we present a high-throughput approach to the exploration of such parameter sets, leveraging recent advances in stream processing hardware (high-end NVIDIA graphic cards and the PlayStation 3's IBM Cell Processor). In analogy to high-throughput screening approaches in molecular biology and genetics, we explored thousands of potential network architectures and parameter instantiations, screening those that show promising object recognition performance for further analysis. We show that this approach can yield significant, reproducible gains in performance across an array of basic object recognition tasks, consistently outperforming a variety of state-of-the-art purpose-built vision systems from the literature. As the scale of available computational power continues to expand, we argue that this approach has the potential to greatly accelerate progress in both artificial vision and our understanding of the computational underpinning of biological vision.

Progress in understanding the brain mechanisms underlying vision requires the construction of computational models that not only emulate the brain's anatomy and physiology, but ultimately match its performance on visual tasks. In recent years, "natural" images have become popular in the study of vision and have been used to show apparently impressive progress in building such models. Here, we challenge the use of uncontrolled "natural" images in guiding that progress. In particular, we show that a simple V1-like model--a neuroscientist's "null" model, which should perform poorly at real-world visual object recognition tasks--outperforms state-of-the-art object recognition systems (biologically inspired and otherwise) on a standard, ostensibly natural image recognition test. As a counterpoint, we designed a "simpler" recognition test to better span the real-world variation in object pose, position, and scale, and we show that this test correctly exposes the inadequacy of the V1-like model. Taken together, these results demonstrate that tests based on uncontrolled natural images can be seriously misleading, potentially guiding progress in the wrong direction. Instead, we reexamine what it means for images to be natural and argue for a renewed focus on the core problem of object recognition--real-world image variation.