Selected Publications

• Core dimensions of human material perception

  

  Authors:

  Schmidt, Hebart, Schmid and Fleming

  Journal:

  Proceedings of the National Academy of Sciences

  Summary:

  We use a large-scale data-driven approach to uncover the core latent dimensions in our mental representation of materials.

  Citation:

  Schmidt, F., Hebart, M. N., Schmid, A. C., & Fleming, R. W. (2025). Core dimensions of human material perception. Proceedings of the National Academy of Sciences, 122(10), e2417202122.

  Stimuli

  Stimuli

  Data

  Data

  
  

• Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

  

  Authors:

  Hartmann, Maiello, Rothkopf & Fleming

  Journal:

  Jove

  Summary:

  We demonstrate a workflow for estimating full contact regions between the human hand and objects as participants grasp them naturally, by measuring intersections between a mesh reconstruction of the hand and a virtual mesh of the object.

  Citation:

  Hartmann, F., Maiello, G., Rothkopf, C. A., Fleming, R. W. Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping. J. Vis. Exp. (194), e64877, https://doi.org/10.3791/64877

  Ressources

  Download data & code

  
  

• Mental object rotation based on two-dimensional visual representations

  

  Authors:

  Stewart, Hartmann, Morgenstern, Storrs, Maiello and Fleming

  Journal:

  Current Biology

  Summary:

  Mental comparisons of 3D object viewpoints have been assumed to occur on 3D representations. We show that non-uniformities in object viewpoint comparisons are captured by a 2D model, suggesting that such computations rely on a ‘mental rendering’ of the simulated proximal viewpoint transformation.

  Citation:

  Stewart, E.E.M., Hartmann, F.T., Morgenstern, Y., Storrs, K.R., Maiello, G., and Fleming, R.W. (in press). Mental object rotation based on two-dimensional visual representations, Current Biology.

  Ressources

  Data

  
  

• One shot generalization in humans revealed through a drawing task

  

  Authors:

  Tiedemann, Morgenstern, Schmidt & Fleming

  Journal:

  eLife

  Summary:

  We analyze how humans are able to create a whole category of objects inspired by justa single exemplar object. We found that observers alter the part-structure of the exemplar in systematic ways, with certain parts beeing deemed as more important than others, to create a novel member of that category.

  Citation:

  Tiedemann, H., Morgenstern, Y., Schmidt, F. & Fleming, R. W. (2022). One shot generalization in humans revealed through a drawing task. eLife, doi:https://doi.org/10.7554/eLife.75485

  Ressources

  Data

  
  

• Unsupervised learning predicts human perception and misperception of gloss.

  

  Authors:

  Storrs, Anderson & Fleming

  Journal:

  Nature Human Behavior

  Summary:

  We find that unsupervised neural networks trained to compress and predict images of surfaces spontaneously learn to organize images by their gloss, lighting and other world factors. The networks correctly predict both successes and striking failures of human gloss constancy, where other models fail.

  Citation:

  Storrs, KR, Anderson, BL & RW Fleming (2021). Unsupervised learning predicts human perception and misperception of gloss. Nature Human Behavior https://doi.org/10.1038/s41562-021-01097-6

  Ressources

  Data

  
  

• The Veiled Virgin illustrates visual segmentation of shape by cause.

  

  Authors:

  Phillips & Fleming

  Journal:

  Proceedings of the National Academy of Sciences

  Summary:

  Inspired by a famous statue depicting the Virgin Mary wearing a transparent veil, we investigated how observers visually segment 3D shapes into multiple 'causal layers'.

  Citation:

  Phillips F* & RW Fleming* (2020). The Veiled Virgin illustrates visual segmentation of shape by cause. Proceedings of the National Academy of Sciences 201917565; DOI: 10.1073/pnas.1917565117

  Ressources

  Data

  
  

• Learning to See Stuff

  

  Authors:

  Fleming & Storrs

  Journal:

  Current Opinion in Behavioral Sciences

  Summary:

  An introduction to unsupervised deep learning approaches to vision, which advocates for such 'statistical appearance models' as being more appropriate ways to think about material perception than classic 'inverse optics' frameworks.

  Citation:

  Fleming RW & KR Storrs (2019) Learning to See Stuff. Current Opinion in Behavioral Sciences. 30: 100–108. https://doi.org/10.1016/j.cobeha.2019.07.004

  
  

• Visual Features in the Perception of Liquids

  

  Authors:

  Van Assen, Barla, Fleming

  Journal:

  Current Biology

  Summary:

  Here, using state-of-the-art fluid simulations, we investigate the cues underlying the perception of liquid viscosity. We suggest that midlevel visual features describing a liquids shape and motion play a key role in viscosity perception Citation:

  Citation:

  Van Assen, JJ, Barla P & Fleming, RW (2018). Visual Features in the Perception of Liquids. Current Biology 28(3), 452–458. https://doi.org/10.1016/j.cub.2017.12.037

  Ressources

  Data

  
  

• Material Perception

  

  Authors:

  Fleming

  Journal:

  Annual Reviews of Vision Science

  Summary:

  This paper reviews recent work on material perception and scrutinizes some basic assumptions of mid-level vision to propose an alternative approach to how material properties might be estimated and represented.

  Citation:

  Fleming RW (2017). Material Perception. Annual Reviews of Vision Science 3(1). 365-388. doi: 10.1146/annurev-vision-102016-061429

  
  

• Flow-Guided Warping for Image-Based Shape Manipulation

  

  Authors:

  Vergne, Barla, Bonneau, Fleming

  Journal:

  ACM Transactions on Graphics

  Summary:

  We present an interactive method that manipulates perceived object shape using a spatial warping technique to exaggerate orientation patterns in the image that are strongly correlated to surface curvature. Our algorithm produces convincing shape manipulation results on synthetic images and photographs, for various materials and lighting environments.

  Citation:

  Vergne, R., Barla P., Bonneau, G.-P. and R. W. Fleming (2016). Flow-Guided Warping for Image-Based Shape Manipulation. ACM Transactions on Graphics (Proceedings of SIGGRAPH 2016), 35(4): 93. DOI: http://dx.doi.org/10.1145/2897824.2925937.

  
  

• 'Proto-rivalry': how the binocular brain identifies gloss

  

  Authors:

  Muryy, Fleming, Welchman

  Journal:

  Proceedings of the Royal Society of London

  Summary:

  Here, we show that the brain exploits some surprisingly simple binocular fusion signals to identify glossy surfaces. We show that the widely held concept of 'binocular luster' is wrong. Instead it is the partial fusion that occurs when binocular matches stray from epipolar geometry that drives binocular gloss perception.

  Citation:

  Muryy, A.A., R.W. Fleming and A.E. Welchman (2016). 'Proto-rivalry': how the binocular brain identifies gloss. Proceedings of the Royal Society of London (B), 283: 20160383. doi: 10.1098/rspb.2016.0383

  
  

• Visual Perception of Materials and their Properties

  

  Authors:

  Fleming

  Journal:

  Vision Research

  Summary:

  Here we review the perception of materials and propose that the visual system uses generative statisical models of material appearance to represent and distinguish between different materials.

  Citation:

  Fleming, R.W. (2014). Visual Perception of Materials and their Properties. Vision Research, 94, 62-75 https://doi.org/10.1016/j.visres.2013.11.004

  
  

• Specular reflections and the estimation of shape from binocular disparity

  

  Authors:

  Muryy, Welchman, Blake, Fleming

  Journal:

  Proceedings of the National Academy of Sciences

  Summary:

  We studied binocular 3D shape perception for ideal mirrored surfaces, finding that far from 'knowing the physics' of specular reflection, they make substantial errors. The responses are consistent with a process that identifies reliable local disparity signals and interpolates between them.

  Citation:

  Muryy, A., Welchman, A.E., Blake, A. and R.W. Fleming (2013). Specular reflections and the estimation of shape from binocular disparity. Proceedings of the National Academy of Sciences, 110(6): 2413-2418. https://doi.org/10.1073/pnas.1212417110

  
  

• Estimation of 3D shape from image orientations

  

  Authors:

  Fleming, Holtmann-Rice, Bülthoff

  Journal:

  Proceedings of the National Academy of Sciences

  Summary:

  Here we show that the perception of 3D shape from surface texture patterns is based primarily on specific 2D orientation signals. Perceptually adapting orientation detectors causes unoriented random noise to look like specific 3D shapes, leading to striking illusions.

  Citation:

  Fleming, R. W., Holtmann-Rice, D. and H. H. Bülthoff (2011). Estimation of 3D shape from image orientations. Proceedings of the National Academy of Sciences, 108(51) 20438-20443. https://doi.org/10.1073/pnas.1114619109

  
  

• Visual Motion and the Perception of Surface Material

  

  Authors:

  Doerschner, Fleming, Yilmaz, Schrater, Hartung, Kersten

  Journal:

  Current Biology

  Summary:

  In this paper we show the central role that optical flow plays in the perception of surface glossiness. We show that sticking reflections onto a surface makes it look matte, and identify specific optic flow characteristics that can predict surface material illusions.

  Citation:

  Doerschner, K. Fleming, R.W., Yilmaz, O., Schrater, P.R., Hartung, B. and Kersten, D. (2011). Visual Motion and the Perception of Surface Material. Current Biology, 21(23): 1-7. https://doi.org/10.1016/j.cub.2011.10.036

  Ressources

  Data