Recurrent neural networks can explain flexible trading of speed and accuracy in biological vision
Open accessThe DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index.
Abstract
Abstract 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. Author summary Deep neural networks provide the best current models of biological vision and achieve the highest performance in computer vision. Inspired by the primate brain, these models transform the image signals through a sequence of stages, leading to recognition. Unlike brains in which outputs of a given computation are fed back into the same computation, these models do not process signals recurrently. The ability to recycle limited neural resources by processing information recurrently could explain the accuracy and flexibility of biological visual systems, which computer vision systems cannot yet match. Here we report that recurrent processing can improve recognition performance compared to similarly complex feedforward networks. Recurrent processing also enabled models to behave more flexibly and trade off speed for accuracy. Like humans, the recurrent network models can compute longer when an object is hard to recognise, which boosts their accuracy. The model’s recognition times predicted human recognition times for the same images. The performance and flexibility of recurrent neural network models illustrates that modeling biological vision can help us improve computer vision.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing
Benjamini, Y., & Hochberg, Y. (1995) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1995.tb02031.x · Cited by 109,482
MobileNetV2: Inverted Residuals and Linear Bottlenecks
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00474 · Cited by 28,755
Xception: Deep Learning with Depthwise Separable Convolutions
Chollet, F. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.195 · Cited by 20,380
Learning Transferable Architectures for Scalable Image Recognition
Zoph, B., Vasudevan, V., Shlens, J., & Le, Q. V. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00907 · Cited by 7,831
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
Iandola, F. N., Han, S., Moskewicz, M. W., Ashraf, K., Dally, W. J., & Keutzer, K. (2016) · arXiv · doi:10.48550/arxiv.1602.07360 · Cited by 6,201
Rao, R. P. N., & Ballard, D. H. (1999) · Nature Neuroscience · doi:10.1038/4580 · Cited by 5,483 · cited 1×
“Such objectives include image classification performance [58], predictive coding [59], temporal stability [60, 61], and sparsity [56].”
BackgroundNote on the Sampling Error of the Difference Between Correlated Proportions or Percentages
McNemar, Q. (1947) · Psychometrika · doi:10.1007/bf02295996 · Cited by 4,754
Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms
Dietterich, T. G. (1998) · Neural Computation · doi:10.1162/089976698300017197 · Cited by 3,840 · cited 1×
“Pairwise McNemar tests [40, 41] showed all differences in model performance to be significant (p� 0.”
BackgroundThe Neural Basis of Decision Making
Gold, J. I., & Shadlen, M. N. (2007) · Annual Review of Neuroscience · doi:10.1146/annurev.neuro.29.051605.113038 · Cited by 3,764 · cited 5×
“If a stimulus falls far away from the decision boundary on the correct side in the multivariate response space, then the evidence will accumulate rapidly [33].”
BackgroundMethodInfluentialHierarchical models of object recognition in cortex
Riesenhuber, M., & Poggio, T. (1999) · Nature Neuroscience · doi:10.1038/14819 · Cited by 3,309 · cited 1×
“Neural networks have a long history as models of biological vision [1–3] and the recent 2 success of deep neural networks (DNNs) in computer vision has led to a renewed 3 interest in neural network models within neuroscience [4–6].”
BackgroundPerformance-optimized hierarchical models predict neural responses in higher visual cortex
Yamins, D. L. K., Hong, H., Cadieu, C. F., Solomon, E. A., Seibert, D., & DiCarlo, J. J. (2014) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1403112111 · Cited by 1,924
How does the brain solve visual object recognition?
DiCarlo, J. J., Zoccolan, D., & Rust, N. C. (2012) · Neuron · doi:10.1016/j.neuron.2012.01.010 · Cited by 1,790
Using goal-driven deep learning models to understand sensory cortex
Yamins, D. L. K., & DiCarlo, J. J. (2016) · Nature Neuroscience · doi:10.1038/nn.4244 · Cited by 1,621
Sparse coding of sensory inputs
Olshausen, B. A., & Field, D. J. (2004) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2004.07.007 · Cited by 1,404
Speed-accuracy tradeoff and information processing dynamics
Wickelgren, W. A. (1977) · Acta Psychologica · doi:10.1016/0001-6918(77)90012-9 · Cited by 1,304 · cited 4×
“We further modelled threshold-based decision making in the rCNNs, varying the threshold to control the SAT [34], and compared reaction times to different images between rCNNs and human observers.”
BackgroundMethodInfluentialDeep supervised, but not unsupervised, models may explain IT cortical representation
Khaligh-Razavi, S.-M., & Kriegeskorte, N. (2014) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1003915 · Cited by 1,204
Self-citationGuclu, U., & van Gerven, M. A. J. (2015) · Journal of Neuroscience · doi:10.1523/jneurosci.5023-14.2015 · Cited by 961
Cadieu, C. F., Hong, H., Yamins, D. L. K., Pinto, N., Ardila, D., Solomon, E. A., Majaj, N. J., & DiCarlo, J. J. (2014) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1003963 · Cited by 824
Anatomy of hierarchy: Feedforward and feedback pathways in macaque visual cortex
Markov, N. T., Vezoli, J., Chameau, P., Falchier, A., Quilodran, R., Huissoud, C., Lamy, C., Misery, P., Giroud, P., Ullman, S., et al. (2014) · Journal of Comparative Neurology · doi:10.1002/cne.23458 · Cited by 759
Freiwald, W. A., & Tsao, D. Y. (2010) · Science · doi:10.1126/science.1194908 · Cited by 678
Kar, K., Kubilius, J., Schmidt, K., Issa, E. B., & DiCarlo, J. J. (2019) · Nature Neuroscience · doi:10.1038/s41593-019-0392-5 · Cited by 496 · cited 3×
“For some images, fast feedforward computations may be sufficient [24].”
BackgroundRecurrence is required to capture the representational dynamics of the human visual system
Kietzmann, T. C., Spoerer, C. J., Sörensen, L. K. A., Cichy, R. M., Hauk, O., & Kriegeskorte, N. (2019) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1905544116 · Cited by 406
Self-citationDeep neural networks in computational neuroscience
Kietzmann, T. C., McClure, P., & Kriegeskorte, N. (2019) · Oxford Research Encyclopedia of Neuroscience · doi:10.1093/acrefore/9780190264086.013.46 · Cited by 186
Self-citation(2017) · doi:10.3389/fncom.2016.00131 · Cited by 59
Self-citation
Cited by
The connectome of an insect brain
Winding, M., Pedigo, B. D., Barnes, C. L., Patsolic, H. G., Park, Y., Kazimiers, T., Fushiki, A., Andrade, I. V., Khandelwal, A., Valdes-Aleman, J., et al. (2023) · Science · doi:10.1126/science.add9330 · Cited by 435
Recurrence is required to capture the representational dynamics of the human visual system
Kietzmann, T. C., Spoerer, C. J., Sörensen, L. K. A., Cichy, R. M., Hauk, O., & Kriegeskorte, N. (2019) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1905544116 · Cited by 406
Self-citationThe neuroconnectionist research programme
Doerig, A., Sommers, R. P., Seeliger, K., Richards, B., Ismael, J., Lindsay, G. W., Kording, K. P., Konkle, T., van Gerven, M. A. J., Kriegeskorte, N., et al. (2023) · Nature Reviews Neuroscience · doi:10.1038/s41583-023-00705-w · Cited by 233
Self-citationAn ecologically motivated image dataset for deep learning yields better models of human vision
Mehrer, J., Spoerer, C. J., Jones, E. C., Kriegeskorte, N., & Kietzmann, T. C. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2011417118 · Cited by 140
Self-citationIndividual differences among deep neural network models
Mehrer, J., Spoerer, C. J., Kriegeskorte, N., & Kietzmann, T. C. (2020) · Nature Communications · doi:10.1038/s41467-020-19632-w · Cited by 133
Self-citation
Metadata
- Authors
- Spoerer, C. J., Kietzmann, T. C., Mehrer, J., Charest, I., & Kriegeskorte, N.
- Venue
- PLOS Computational Biology
- Publisher
- Public Library of Science (PLoS)
- Published
- 2020-10-02
- Volume
- 16
- Issue
- 10
- Pages
- e1008215
- ISSN
- 1553-734X (ISSN-L)
- 1553-7358
- OpenAlex
- W3090234134
- Cited by
- 121
- References
- 26
- FWCI
- 7.76
- Type
- journal article
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.