Download - Motion anticipation in the retina
![Page 1: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/1.jpg)
HAL Id: hal-02316888https://hal.inria.fr/hal-02316888
Submitted on 15 Oct 2019
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
Motion anticipation in the retinaBruno Cessac, Selma Souihel
To cite this version:Bruno Cessac, Selma Souihel. Motion anticipation in the retina. NeuroSTIC 2019 - 7e édition desjournées NeuroSTIC, Oct 2019, Sophia-Antipolis, France. �hal-02316888�
![Page 2: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/2.jpg)
Motion anticipation in the retina
Bruno Cessac, Selma Souihel
Biovision INRIA team
![Page 3: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/3.jpg)
Motion anticipation in the retina
Bruno Cessac, Selma Souihel
Biovision INRIA team
![Page 4: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/4.jpg)
Motion anticipation in the retina
Bruno Cessac, Selma Souihel
In collaboration with :
Frédéric ChavaneSandrine Chemla
Olivier MarreMatteo Di VoloAlain Destexhe
Biovision INRIA team
![Page 5: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/5.jpg)
4
The visual flow
Source : Wikipedia
![Page 6: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/6.jpg)
5
The visual flow
Source : Wikipedia
![Page 7: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/7.jpg)
6
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 8: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/8.jpg)
7
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 9: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/9.jpg)
8
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 10: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/10.jpg)
9
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 11: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/11.jpg)
10
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 12: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/12.jpg)
11
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 13: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/13.jpg)
12
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 14: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/14.jpg)
13
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
Too slow !
![Page 15: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/15.jpg)
14
Visual Anticipation
Source : Benvenutti et al. 2015
![Page 16: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/16.jpg)
15
Visual Anticipation
Source : Benvenutti et al. 2015
Anticipation is carried out by the primary visual cortex (V1) through an activation wave
![Page 17: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/17.jpg)
16
Visual Anticipation
Source :Berry et al.1999
Anticipation also takes place in the retina
![Page 18: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/18.jpg)
17
Visual Anticipation
What are the respective mechanisms underlying retinaland cortical anticipation?
TrajectoryTrajectory
![Page 19: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/19.jpg)
18
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
The retina is NOT a camera
• High transduction rate : 1 photon cantrigger a membrane voltage variation of~1 mV
• Able to detect approaching motion
• Able to detect differential motion
• Sensitive to « surprise » in a visualscene
• Able to perform motion anticipation
![Page 20: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/20.jpg)
19
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 21: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/21.jpg)
20
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion« Analogic computing »
No spikes (except in GCs)Low energy consumption
Dedicated circuitsSmall number of neurons
Specialized synapses
![Page 22: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/22.jpg)
21
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 23: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/23.jpg)
22
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 24: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/24.jpg)
23
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
Horizontal cells
![Page 25: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/25.jpg)
24
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
Horizontal cells
Amacrine cells
![Page 26: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/26.jpg)
25
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
![Page 27: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/27.jpg)
26
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion
Gap junctions
![Page 28: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/28.jpg)
27
The visual flow
Source : Wikipedia
Decoding spike trains
Encoding motion« Analogic computing »
No spikes (except in GCs)Low energy consumption
Dedicated circuitsSmall number of neurons
Specialized synapses
![Page 29: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/29.jpg)
28
Which generic computational paradigms are atwork in the retina ?
![Page 30: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/30.jpg)
29
Visual Anticipation
![Page 31: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/31.jpg)
30
Visual Anticipation
Which animal ?
![Page 32: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/32.jpg)
31
Visual Anticipation
Which animal ?
![Page 33: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/33.jpg)
32
Visual Anticipation
Which animal ?
![Page 34: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/34.jpg)
33
Visual Anticipation
Which animal ?
![Page 35: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/35.jpg)
34
Visual Anticipation
Which animal ?
![Page 36: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/36.jpg)
35
Visual Anticipation
Which animal ?
![Page 37: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/37.jpg)
36
Visual Anticipation
Developping a retino-cortical model of anticipation soas to
understand / propose
possible generic mechanisms for anticipation in theretina and in the cortex.
![Page 38: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/38.jpg)
37
Anticipation in the retina
![Page 39: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/39.jpg)
38
The Hubel-Wiesel view of vision
Ganglion cells
Nobel prize 1981
Ganglion cells response is the convolution of the stimulus with a spatio-temporalreceptive field followed by a non linearity
Ganglion cells are independent encoders
![Page 40: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/40.jpg)
39
The Hubel-Wiesel view of vision
Source : Berry et al. 1999
Ganglion cells
Nobel prize 1981
![Page 41: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/41.jpg)
40
Gain control (Berry et al, 1999, Chen et al. 2013)
Building a 2D retina model for motionanticipation
![Page 42: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/42.jpg)
41
Building a 2D retina model for motionanticipation
Gain control (Berry et al, 1999, Chen et al. 2013)
![Page 43: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/43.jpg)
42
Building a 2D retina model for motionanticipation
Gain control (Chen et al. 2013)
![Page 44: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/44.jpg)
43
Building a 2D retina model for motionanticipation
![Page 45: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/45.jpg)
44
1D results : smooth motion anticipationwith gain control
Bipolar layer Ganglionlayer
![Page 46: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/46.jpg)
45
1D results : smooth motion anticipationwith gain control
Anticipation variability with stimulusparameters
![Page 47: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/47.jpg)
46
Building a 2D retina model for motionanticipation
Ganglion cells are independent encoders
![Page 48: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/48.jpg)
47
Building a 2D retina model for motionanticipation
Ganglion cells are not independent encoders
Gap junctions connectivity
![Page 49: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/49.jpg)
48
Building a 2D retina model for motionanticipation
Gap junctions connectivity
![Page 50: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/50.jpg)
49
Building a 2D retina model for motionanticipation
Gap junctions connectivity
![Page 51: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/51.jpg)
50
Building a 2D retina model for motionanticipation
Gap junctions connectivity
![Page 52: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/52.jpg)
51
Building a 2D retina model for motionanticipation
Diffusive wave of activity ahead of the motion
Gap junctions connectivity
![Page 53: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/53.jpg)
52
Smooth motion anticipation with gapjunctions
![Page 54: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/54.jpg)
53
6
Low gap junction conductance
Smooth motion anticipation with gapjunctions
![Page 55: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/55.jpg)
54
6
Low gap junction conductance High gap junction conductance
Smooth motion anticipation with gapjunctions
![Page 56: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/56.jpg)
55
6
Smooth motion anticipation with gapjunctions
![Page 57: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/57.jpg)
56
Anticipation variability with stimulusparameters
Smooth motion anticipation with gapjunctions
![Page 58: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/58.jpg)
57
Building a 2D retina model for motionanticipation
![Page 59: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/59.jpg)
58
Building a 2D retina model for motionanticipation
Ganglion cells are not independent encoders
Amacrine cells connectivity
![Page 60: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/60.jpg)
59
Amacrine cells connectivity
Building a 2D retina model for motionanticipation
![Page 61: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/61.jpg)
60
Amacrine cells connectivity
Building a 2D retina model for motionanticipation
![Page 62: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/62.jpg)
61
Amacrine cells connectivity
● The circuitry involves amacrine cells connectivity upstream of ganglion cells
Building a 2D retina model for motionanticipation
● A class of RGCs are selective to differential motion
![Page 63: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/63.jpg)
62
Amacrine cells connectivity
Building a 2D retina model for motionanticipation
![Page 64: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/64.jpg)
63
Amacrine cells connectivity
Building a 2D retina model for motionanticipation
![Page 65: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/65.jpg)
64
Amacrine cells connectivity
Diffusive wave of activityahead of the bar
Building a 2D retina model for motionanticipation
![Page 66: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/66.jpg)
65
Building a 2D retina model for motionanticipation
![Page 67: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/67.jpg)
66
Building a 2D retina model for motionanticipation
![Page 68: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/68.jpg)
67
Building a 2D retina model for motionanticipation
![Page 69: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/69.jpg)
68
Building a 2D retina model for motionanticipation
![Page 70: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/70.jpg)
69
Building a 2D retina model for motionanticipation
Spatial profiles w=1
x
![Page 71: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/71.jpg)
70
Building a 2D retina model for motionanticipation
Spatial profiles w=3
x
![Page 72: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/72.jpg)
71
Building a 2D retina model for motionanticipation
Spatial profiles w=5
x
![Page 73: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/73.jpg)
72
Building a 2D retina model for motionanticipation
Temporal profile of the middle cell
![Page 74: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/74.jpg)
73
1D results : smooth motion anticipationwith amacrine connectivity
Bipolar layer Ganglion layer
![Page 75: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/75.jpg)
74
1D results : smooth motion anticipationwith amacrine connectivity
Anticipation variability with stimulusparameters
![Page 76: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/76.jpg)
75
Comparing the performance of the three layers
![Page 77: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/77.jpg)
76
Suggesting new experiments : 2D results
1) Angular anticipation
Stimulus
t = 0 ms 100 200 ms 300 ms 400 ms 500 ms 600 ms 700 ms
Bipolar linearresponse
Bipolar gainresponse
Ganglion linearresponse
Ganglion gainresponse
A)
B) C)
![Page 78: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/78.jpg)
77
Suggesting new experiments : 2D results
1) Angular anticipation
![Page 79: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/79.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
![Page 80: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/80.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
● A large part of the « computation » is made by synapses.
![Page 81: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/81.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
● A large part of the « computation » is made by synapses.
● Starting from the retina architecture one can extract circuits solving « tasks » such asmotion anticipation.
![Page 82: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/82.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
● A large part of the « computation » is made by synapses.
● Starting from the retina architecture one can extract circuits solving « tasks » such asmotion anticipation.
● The role of gain control and local synaptic balance between excitation-inhibition hasbeen experimentally shown to play a central rôle in anticipation (Berry et al, 1999 ; Chenat al, 2013 ; Johnston-Lagando, 2015).
![Page 83: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/83.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
● A large part of the « computation » is made by synapses.
● Starting from the retina architecture one can extract circuits solving « tasks » such asmotion anticipation.
● The role of gain control and local synaptic balance between excitation-inhibition hasbeen experimentally shown to play a central rôle in anticipation (Berry et al, 1999 ; Chenat al, 2013 ; Johnston-Lagando, 2015).
● Here we propose that lateral connectivity also plays a role in motion anticipationwhere a wave of activity propagates ahead of the motion.
![Page 84: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/84.jpg)
Conclusions
● The retina is able to encode complex visual scene features, fast and reliably, with avery low energy consumption, without spikes, except at the ganglion cells level.
● A large part of the « computation » is made by synapses.
● Starting from the retina architecture one can extract circuits solving « tasks » such asmotion anticipation.
● The role of gain control and local synaptic balance between excitation-inhibition hasbeen experimentally shown to play a central rôle in anticipation (Berry et al, 1999 ; Chenat al, 2013 ; Johnston-Lagando, 2015).
● Here we propose that lateral connectivity also plays a role in motion anticipationwhere a wave of activity propagates ahead of the motion.
● Useful paradigms for :1) Computer vision ?2) Retinal prostheses ?
![Page 85: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/85.jpg)
84
Anticipation in V1
![Page 86: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/86.jpg)
85
Anticipation in V1
![Page 87: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/87.jpg)
86
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
![Page 88: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/88.jpg)
87
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
![Page 89: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/89.jpg)
88
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
Affords a retino thalamic input
![Page 90: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/90.jpg)
89
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
![Page 91: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/91.jpg)
90
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
![Page 92: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/92.jpg)
91
A mean field model to reproduce VSDIrecordings Zerlaut et al 2016
Chemla et al 2018
![Page 93: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/93.jpg)
Response of the cortical model to a LNretina drive
![Page 94: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/94.jpg)
Response of the cortical model to a retinadrive with gain control
![Page 95: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/95.jpg)
Anticipation in the cortex : VSDI dataanalysis (Data courtesy of F.
Chavane et S. Chemla)
![Page 96: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/96.jpg)
Comparing simulation results to VSDIrecordings
Cortex experimentalrecordings
Simulation resultsResponse to an LNmodel of the retina
Simulation resultsResponse to a gaincontrol model of theretina
![Page 97: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/97.jpg)
Conclusions
● We developped a 2D retina with three ganglion cell layers,implementing gain control and connectivity.
● We use the output of our model as an input to a mean field model ofV1, and were able to reproduce anticipation as observed in VSDI
![Page 98: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/98.jpg)
Conclusions
● How to improve object identification ● 1) exploring the model's parameters and
● 2) using connectivity ?
● Is our model able to anticipate more complex trajectories, withaccelerations for instance ?
● How to calibrate connectivity using biology ?
● How does anticipation affect higher order correlations ?
● Would it be possible to design psycho-physical tests clearly showingthe role of the retina in visual anticipation ?
![Page 99: Motion anticipation in the retina](https://reader030.vdocuments.us/reader030/viewer/2022012810/61c0844f0311c209f333b808/html5/thumbnails/99.jpg)
Thank you for your attention !