Corpus ID: 139104798

Capturing human categorization of natural images at scale by combining deep networks and cognitive models

@article{Battleday2019CapturingHC,
  title={Capturing human categorization of natural images at scale by combining deep networks and cognitive models},
  author={Ruairidh M. Battleday and Joshua C. Peterson and Thomas L. Griffiths},
  journal={ArXiv},
  year={2019},
  volume={abs/1904.12690}
}
Human categorization is one of the most important and successful targets of cognitive modeling in psychology, yet decades of development and assessment of competing models have been contingent on small sets of simple, artificial experimental stimuli. Here we extend this modeling paradigm to the domain of natural images, revealing the crucial role that stimulus representation plays in categorization and its implications for conclusions about how people form categories. Applying psychological… Expand
Training Deep Networks to Construct a Psychological Feature Space for a Natural-Object Category Domain
TLDR
A novel integration in which MDS representations are used to train deep convolutional neural networks (CNNs) to automatically derive psychological representations for unlimited numbers of natural stimuli to extend traditional psychological-scaling and category-learning models to the complex, high-dimensional domains that exist in the natural world. Expand
Integrating a Gaussian Classifier into a CNN
The assignment of natural images to categories is an important problem both in machine learning and in the study of human cognition. Visual categorization models in psychology examine strategies ofExpand
Combining Convolutional Neural Networks and Cognitive Models to Predict Novel Object Recognition in Humans
TLDR
This work uses CNN object representations, built up over a network hierarchy from object images, to drive a cognitive model of decision making, the Linear Ballistic Accumulator (LBA), to predict a range of performance in a visual matching task with novel objects. Expand
End-to-end Deep Prototype and Exemplar Models for Predicting Human Behavior
TLDR
This work extends classic prototype and exemplar models to learn both stimulus and category representations jointly from raw input and finds that their cognitively inspired counterparts both provide better intrinsic fit to human behavior and improve ground-truth classification. Expand
A neural network walks into a lab: towards using deep nets as models for human behavior
TLDR
It is argued that methods for assessing the goodness of fit between DNN models and human behavior have to date been impoverished, and cognitive science might have to start using more complex tasks, but doing so might be beneficial for DNN-independent reasons as well. Expand
Contrasting Exemplar and Prototype Models in a Natural-Science Category Domain
TLDR
The results of the qualitative and quantitative model comparisons point toward the exemplar model as providing a better account of the observed results. Expand
THINGSvision: a Python toolbox for streamlining the extraction of activations from deep neural networks
TLDR
ThINGSvision enables researchers across diverse fields to extract features in a streamlined manner for their custom image dataset, thereby improving the ease of relating DNNs, brain activity, and behavior, and improving the reproducibility of findings in these research fields. Expand
Representing absence of evidence: why algorithms and representations matter in models of language and cognition
TLDR
Theories of language and cognition develop iteratively from ideas, experiments and models, and computational models play a critical role in this development. Expand
Controversial stimuli: Pitting neural networks against each other as models of human cognition
TLDR
This work synthesized controversial stimuli: images for which different models produce distinct responses, and found that deep neural networks, which model the distribution of images, performed better than purely discriminative DNNs, which learn only to map images to labels. Expand

References

SHOWING 1-10 OF 40 REFERENCES
Evaluating (and Improving) the Correspondence Between Deep Neural Networks and Human Representations
TLDR
It is found that state-of-the-art object classification networks provide surprisingly accurate predictions of human similarity judgments for natural images, but they fail to capture some of the structure represented by people. Expand
Using Deep-Learning Representations of Complex Natural Stimuli as Input to Psychological Models of Classification
TLDR
This work trains an ensemble of convolutional neural networks to produce the multidimensional scaling coordinates of images of rocks and shows that not only are the CNNs able to predict the MDS coordinates of a held-out test set of rocks, but that the CNN-derived representations can be used in combination with a formal psychological model to predict human categorization behavior on a completely new set of Rocks. Expand
Deep Neural Networks Predict Category Typicality Ratings for Images
TLDR
It is found that deep convolutional networks trained for classification have substantial predictive power, unlike simpler features computed from the same massive dataset, showing how typicality might emerge as a byproduct of a complex model trained to maximize classification performance. Expand
Pixels to Voxels: Modeling Visual Representation in the Human Brain
TLDR
The fit models provide a new platform for exploring the functional principles of human vision, and they show that modern methods of computer vision and machine learning provide important tools for characterizing brain function. Expand
Learning Multiple Layers of Features from Tiny Images
TLDR
It is shown how to train a multi-layer generative model that learns to extract meaningful features which resemble those found in the human visual cortex, using a novel parallelization algorithm to distribute the work among multiple machines connected on a network. Expand
Attention, similarity, and the identification-categorization relationship.
  • R. Nosofsky
  • Medicine
  • Journal of experimental psychology. General
  • 1986
TLDR
A unified quantitative approach to modeling subjects' identification and categorization of multidimensional perceptual stimuli is proposed and tested and some support was gained for the hypothesis that subjects distribute attention among component dimensions so as to optimize categorization performance. Expand
Human Object-Similarity Judgments Reflect and Transcend the Primate-IT Object Representation
TLDR
It is shown that objects that elicit similar activity patterns in human IT (hIT) tend to be judged as similar by humans, and IT was more similar to monkey IT than to human judgments. Expand
The development of features in object concepts.
TLDR
This work presents a theory in which people create features to subserve the representation and categorization of objects, and presents evidence of flexible perceptual changes arising from category learning and theoretical arguments for the importance of this flexibility. Expand
Handbook of Categorization in Cognitive Science
The discussion of whether emotions are categorical in nature has loomed large ever since Darwin (1872) posited an evolutionary base for human emotion expressions. This discussion has traditionallyExpand
On Learning Natural-Science Categories That Violate the Family-Resemblance Principle
TLDR
This investigation documented an example of an important exception to this principle: Results of a multidimensional-scaling study of igneous, metamorphic, and sedimentary rocks suggested that the structure of these categories is disorganized and dispersed. Expand
...
1
2
3
4
...