In this work we tested a relatively small number ofPathTracker versions. We mostly focused on small variations to the number of distractors and video length, but in future work we hope to incorporate other variations like speed and velocity manipulations, and generalization across temporalvariations[1]. One potential issue is determining when a visual system should rely on appearance-based vs. appearance-free features for tracking. Our solution is two-pronged and potentially insufficient. The first strategy is for top-down feedback from the TransT into the InT,which we aligns tracks between the two models.

Themanysuccesses ofdeepneural networks(DNNs) overthepastdecade have largely been driven by computational scale rather than insights from biological intelligence.

The many successes of deep neural networks (DNNs) over the past decade have largely been driven by computational scale rather than insights from biological intelligence. Here, we explore if these trends have also carried concomitant improvements in explaining the visual strategies humans rely on for object recognition. We do this by comparing two related but distinct properties of visual strategies in humans and DNNs: where they believe important visual features are in images and how they use those features to categorize objects. Across 84 different DNNs trained on ImageNet and three independent datasets measuring the where and the how of human visual strategies for object recognition on those images, we find a systematic trade-off between DNN categorization accuracy and alignment with human visual strategies for object recognition.

In this work we tested a relatively small number of PathTracker versions. One potential issue is determining when a visual system should rely on appearance-based vs. appearance-free features for tracking. Our solution is two-pronged and potentially insufficient. The first strategy is for top-down feedback from the TransT into the InT, which we aligns tracks between the two models. Additional work is needed to identify better approaches.

The many successes of deep neural networks (DNNs) over the past decade have largely been driven by computational scale rather than insights from biological intelligence. Here, we explore if these trends have also carried concomitant improvements in explaining the visual strategies humans rely on for object recognition. We do this by comparing two related but distinct properties of visual strategies in humans and DNNs: where they believe important visual features are in images and how they use those features to categorize objects. Across 84 different DNNs trained on ImageNet and three independent datasets measuring the where and the how of human visual strategies for object recognition on those images, we find a systematic trade-off between DNN categorization accuracy and alignment with human visual strategies for object recognition. We rectify this growing issue with our neural harmonizer: a general-purpose training routine that both aligns DNN and human visual strategies and improves categorization accuracy.

The many successes of deep neural networks (DNNs) over the past decade have largely been driven by computational scale rather than insights from biological intelligence. Here, we explore if these trends have also carried concomitant improvements in explaining the visual strategies humans rely on for object recognition. We do this by comparing two related but distinct properties of visual strategies in humans and DNNs: where they believe important visual features are in images and how they use those features to categorize objects. Across 84 different DNNs trained on ImageNet and three independent datasets measuring the where and the how of human visual strategies for object recognition on those images, we find a systematic trade-off between DNN categorization accuracy and alignment with human visual strategies for object recognition. State-of-the-art DNNs are progressively becoming less aligned with humans as their accuracy improves. We rectify this growing issue with our neural harmonizer: a general-purpose training routine that both aligns DNN and human visual strategies and improves categorization accuracy. Our work represents the first demonstration that the scaling laws [1-3] that are guiding the design of DNNs today have also produced worse models of human vision.

We present two visual algorithms, called the affine and fractal methods, which each solve a considerable portion of the Raven’s Progressive Matrices (RPM) test. The RPM is considered to be one of the premier psychometric measures of general intelligence. Current computational accounts of the RPM assume that visual test inputs are translated into propositional representations before further reasoning takes place. We propose that visual strategies can also solve RPM problems, in line with behavioral evidence showing that humans do use visual strategies to some extent on the RPM. Our two visual methods currently solve RPM problems at the level of typical 9- to 10-year-olds.