Sparse coding predicts optic flow specificities of zebrafish pretectal neurons

  • Gerrit A. EckeEmail author
  • Sebastian A. Bruijns
  • Johannes Hölscher
  • Fabian A. Mikulasch
  • Thede Witschel
  • Aristides B. Arrenberg
  • Hanspeter A. Mallot
Brain inspired Computing&Machine Learning Applied Research-BISMLARE


Zebrafish pretectal neurons exhibit specificities for large-field optic flow patterns associated with rotatory or translatory body motion. We investigate the hypothesis that these specificities reflect the input statistics of natural optic flow. Realistic motion sequences were generated using computer graphics simulating self-motion in an underwater scene. Local retinal motion was estimated with a motion detector and encoded in four populations of directionally tuned retinal ganglion cells, represented as two signed input variables. This activity was then used as input into one of three learning networks: a sparse coding network (competitive learning), PCA whitening with subsequent sparse coding, and a backpropagation network (supervised learning). All simulations developed specificities for optic flow which are comparable to those found in a neurophysiological study (Kubo et al. in Neuron 81(6):1344–1359, 2016., but relative frequencies of the various neuronal responses were best modeled by the sparse coding approach without whitening. We conclude that the optic flow neurons in the zebrafish pretectum do reflect the optic flow statistics. The predicted vectorial receptive fields show not only typical optic flow fields but also “Gabor” and dipole-shaped patterns that likely reflect difference fields needed for reconstruction by linear superposition.


Optic flow Sparse coding Optimality pretectum Egomotion detection 



This work was carried out at the Department of Biology of the Eberhard-Karls-University, Tübingen, Germany. Additional support was obtained for TW from the Deutsche Forschungsgemeinschaft within the Werner Reichardt Center for Integrative Neuroscience (CIN), Tübingen.

Compliance with ethical standards

Conflict of interest

The authors declared that they have no conflicts of interest to this work.

Supplementary material

521_2019_4500_MOESM1_ESM.pdf (108.3 mb)
Supplementary material 1 (pdf 110938 KB)
521_2019_4500_MOESM2_ESM.pdf (92.1 mb)
Supplementary material 2 (pdf 94286 KB)
521_2019_4500_MOESM3_ESM.pdf (95.4 mb)
Supplementary material 3 (pdf 97708 KB)


  1. 1.
    Antinucci P et al (2016) Neural mechanisms generating orientation selectivity in the retina. Curr Biol 26(14):1802–1815. CrossRefGoogle Scholar
  2. 2.
    Bak-Coleman J, Smith D, Coombs S (2015) Going with, then against the flow: evidence against the optomotor hypothesis of fish rheotaxis. Anim Behav 107:7–17. CrossRefGoogle Scholar
  3. 3.
    Barlow HB (1972) Single units and sensation: A neuron doctrine for perceptual psychology? Perception 1(4):371–394. CrossRefGoogle Scholar
  4. 4.
    Dosovitskiy A et al (2015) FlowNet: learning optical flow with convolutional networks. In: 2015 IEEE international conference on computer vision (ICCV). IEEE, pp 2758–2766.
  5. 5.
    Franz MO, Chahl JS, Krapp HG (2004) Insect-inspired estimation of egomotion. Neural Comput 16(11):2245–2260. CrossRefzbMATHGoogle Scholar
  6. 6.
    Honegger KS, Campbell RAA, Turner GC (2011) Cellular-resolution population imaging reveals robust sparse coding in the drosophila mushroom body. J Neurosci 31(33):11772–11785. CrossRefGoogle Scholar
  7. 7.
    Hyvärinen A, Hurri J, Hoyer PO (2009) Natural image statistics. Springer, London. CrossRefzbMATHGoogle Scholar
  8. 8.
    Hyvärinen A, Oja E (2000) Independent component analysis: algorithms and applications. Neural Netw 13(4):411–430CrossRefGoogle Scholar
  9. 9.
    Ilg E et al (2017) FlowNet 2.0: evolution of optical flow estimation with deep networks. In: 2017 IEEE conference on computer vision and pattern recognition (CVPR). IEEE.
  10. 10.
    Kubo F et al (2014) Functional architecture of an optic flow-responsive area that drives horizontal eye movements in zebrafish. Neuron 81(6):1344–1359. CrossRefGoogle Scholar
  11. 11.
    Marr D (1982) Vision: a computational investigation into the human representation and processing of visual information. W.H. Freeman, New YorkGoogle Scholar
  12. 12.
    Nikolaou N et al (2012) Parametric functional maps of visual inputs to the tectum. Neuron 76(2):317–324. CrossRefGoogle Scholar
  13. 13.
    Olshausen BA, Field DJ (1996) Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 381(6583):607–609. CrossRefGoogle Scholar
  14. 14.
    Olshausen BA, Field DJ (2005) How close are we to understanding V1? Neural Comput 17(8):1665–1699. CrossRefzbMATHGoogle Scholar
  15. 15.
    Olshausen BA, Field DJ (1997) Sparse coding with an overcomplete basis set: a strategy employed by V1? Vis Res 37(23):3311–3325. CrossRefGoogle Scholar
  16. 16.
    Orban GA (2008) Higher order visual processing in macaque extrastriate cortex. Physiol Rev 88(1):59–89. CrossRefGoogle Scholar
  17. 17.
    Papadopoulou M et al (2011) Normalization for sparse encoding of odors by a wide-field interneuron. Science 332(6030):721–725. CrossRefGoogle Scholar
  18. 18.
    Perrone JA (1992) Model for the computation of self-motion in biological systems. J Opt Soc Am A 9(2):177. MathSciNetCrossRefGoogle Scholar
  19. 19.
    Raudies F, Neumann H (2012) A review and evaluation of methods estimating ego-motion. Comput Vis Image Underst 116(5):606–633. CrossRefGoogle Scholar
  20. 20.
    Rozell CJ et al (2008) Sparse coding via thresholding and local competition in neural circuits. Neural Comput 20(10):2526–2563. MathSciNetCrossRefGoogle Scholar
  21. 21.
    Schultz PF et al (2014) Replicating kernels with a short stride allows sparse reconstructions with fewer independent kernels. arXiv preprint arXiv:1406.4205
  22. 22.
    Spence R et al (2007) The behaviour and ecology of the zebrafish, Danio rerio. Biol Rev 83(1):13–34. CrossRefGoogle Scholar
  23. 23.
    Timofte R, Van Gool L (2015) Sparse flow: sparse matching for small to large displacement optical flow. In: 2015 IEEE winter conference on applications of computer vision. IEEE, pp 1100–1106.
  24. 24.
    Verri A, Girosi F, Torre V (1989) Mathematical properties of the two-dimensional motion field: from singular points to motion parameters. JOSA A 6(5):698–712MathSciNetCrossRefGoogle Scholar
  25. 25.
    Vijayanarasimhan S et al (2017) Sfm-net: learning of structure and motion from video. arXiv preprint arXiv:1704.07804
  26. 26.
    Wulff J, Black MJ (2015) Efficient sparse-to-dense optical flow estimation using a learned basis and layers. In: 2015 IEEE conference on computer vision and pattern recognition (CVPR). IEEE, pp 120–130.
  27. 27.
    Zhou T et al (2017) Unsupervised learning of depth and ego-motion from video. In: 2017 IEEE conference on computer vision and pattern recognition (CVPR). IEEE.

Copyright information

© Springer-Verlag London Ltd., part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of BiologyUniversity of TübingenTübingenGermany

Personalised recommendations