A Balanced Comparison of Object Invariances in Monkey IT Neurons

Abstract Our ability to recognize objects across variations in size, position, or rotation is based on invariant object representations in higher visual cortex. However, we know little about how these invariances are related. Are some invariances harder than others? Do some invariances arise faster than others? These comparisons can be made only upon equating image changes across transformations. Here, we targeted invariant neural representations in the monkey inferotemporal (IT) cortex using object images with balanced changes in size, position, and rotation. Across the recorded population, IT neurons generalized across size and position both stronger and faster than to rotations in the image plane as well as in depth. We obtained a similar ordering of invariances in deep neural networks but not in low-level visual representations. Thus, invariant neural representations dynamically evolve in a temporal order reflective of their underlying computational complexity.

• Publication year

  2017

• Venue

  eNeuro

• Publication date

  2017-03-01

• Fields of study

  Biology, Mathematics, Medicine

• Identifiers
  DOI 10.1523/ENEURO.0333-16.2017PMID 28413827PMCID 5390242
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Do Computational Models Differ Systematically from Human Object Perception?

  2016cited by this paper
• Seeing a straight line on a curved surface: decoupling of patterns from surfaces by single IT neurons

  2016cited by this paper
• Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition

  2015cited by this paper
• Dynamics of 3D view invariance in monkey inferotemporal cortex

  2015cited by this paper
• The dynamics of invariant object recognition in the human visual system.

  2014cited by this paper
• Very Deep Convolutional Networks for Large-Scale Image Recognition

  2014cited by this paper
• Performance-optimized hierarchical models predict neural responses in higher visual cortex

  2014cited by this paper
• Deep Supervised, but Not Unsupervised, Models May Explain IT Cortical Representation

  2014cited by this paper
• The Invariance Hypothesis Implies Domain-Specific Regions in Visual Cortex

  2014cited by this paper
• How does the brain solve visual object recognition?

  2012cited by this paper
• Balanced Increases in Selectivity and Tolerance Produce Constant Sparseness along the Ventral Visual Stream

  2012cited by this paper
• Selectivity and Tolerance (“Invariance”) Both Increase as Visual Information Propagates from Cortical Area V4 to IT

  2010cited by this paper
• What response properties do individual neurons need to underlie position and clutter "invariant" object recognition?

  2009cited by this paper
• Why is Real-World Visual Object Recognition Hard?

  2008cited by this paper
• Trade-Off between Object Selectivity and Tolerance in Monkey Inferotemporal Cortex

  2007cited by this paper
• Transformation of shape information in the ventral pathway.

  2007cited by this paper
• Fast Readout of Object Identity from Macaque Inferior Temporal Cortex

  2005cited by this paper
• Viewer-centred and object-centred coding of heads in the macaque temporal cortex

  2004cited by this paper
• How Does Our Visual System Achieve Shift and Size Invariance

  2004cited by this paper
• Empirical evidence for independent dimensions in the visual representation of three-dimensional shape.

  2002cited by this paper
• Recognizing novel three–dimensional objects by summing signals from parts and views

  2002cited by this paper
• Differing views on views: response to Hayward and Tarr (2000).

  2000cited by this paper
• Hierarchical models of object recognition in cortex

  1999cited by this paper
• Responses to contour features in macaque area V4.

  1999cited by this paper
• Three-dimensional object recognition is viewpoint dependent

  1998cited by this paper
• Testing conditions for viewpoint invariance in object recognition.

  1997cited by this paper
• Invariant face and object recognition in the visual system.

  1997cited by this paper
• Inferotemporal cortex and object vision.

  1996cited by this paper
• Size and position invariance of neuronal responses in monkey inferotemporal cortex.

  1995cited by this paper

• Neural representations of perspectival shapes and attentional effects: Evidence from fMRI and MEG.

  2024influential citation
• Unraveling the complexity of rat object vision requires a full convolutional network and beyond

  2024cites this paper
• Relationships between the degrees of freedom in the affine Gaussian derivative model for visual receptive fields and 2-D affine image transformations with application to covariance properties of simple cells in the primary visual cortex

  2024cites this paper
• From intermediate shape-centered representations to the perception of oriented shapes: response to commentaries

  2023cites this paper
• Adversarially trained neural representations may already be as robust as corresponding biological neural representations

  2022cites this paper
• Different patterns of visual processing for shapes in V1 of mouse and monkey

  2022cites this paper
• Invariance of object detection in untrained deep neural networks

  2022influential citation
• Convolutional neural networks do not develop brain-like transformation tolerant visual representations

  2021cites this paper
• Object discrimination performance and dynamics evaluated by inferotemporal cell population activity

  2021cites this paper
• Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future

  2020cites this paper
• Understanding transformation tolerant visual object representations in the human brain and convolutional neural networks

  2020cites this paper
• Stable readout of observed actions from format-dependent activity of monkey’s anterior intraparietal neurons

  2020cites this paper
• Characterisation of nonlinear receptive fields of visual neurons by convolutional neural network

  2019cites this paper
• Multiplicative mixing of object identity and image attributes in single inferior temporal neurons

  2018cites this paper
• Characterization of nonlinear receptive fields of visual neurons by convolutional neural network

  2018cites this paper
• Three-stage processing of category and variation information by entangled interactive mechanisms of peri-occipital and peri-frontal cortices

  2018cites this paper