Biovision team 2 Retina Visual cortex 3 Retina Visual cortex 3 - - PowerPoint PPT Presentation

▶
biovision team 2
SMART_READER_LITE
LIVE PREVIEW

Biovision team 2 Retina Visual cortex 3 Retina Visual cortex 3 - - PowerPoint PPT Presentation

Jan 10, 2023 106 likes •750 views

Biovision team 2 Retina Visual cortex 3 Retina Visual cortex 3 Retina Visual cortex 3 Retina Visual cortex 3 285 millions visually impaired people Retina Visual cortex 3 285 millions visually impaired people Retina Visual cortex

slide-1
SLIDE 1

Biovision team

slide-2
SLIDE 2

2

slide-3
SLIDE 3

Retina Visual cortex

3

slide-4
SLIDE 4

Retina Visual cortex

3

slide-5
SLIDE 5

Retina Visual cortex

3

slide-6
SLIDE 6

Retina Visual cortex

3

slide-7
SLIDE 7

Retina Visual cortex

285 millions visually impaired people

3

slide-8
SLIDE 8

Retina Visual cortex

285 millions visually impaired people

3

New scientific and technological challenges New paradigms to understand vision New technological breakthroughs New technics are emerging to help people use their remaining vision, slow down or even reverse vision loss

slide-9
SLIDE 9

Visual impairment examples

4

From low vision to blindness

slide-10
SLIDE 10

Visual impairment examples

4

Glaucoma (ganglion cells and the optic nerve degenerate) Retinitis Pigmentosa (rods photoreceptors degenerate) Age Related Macular Degeneration (macula cones degenerate) Degenerative myopia (major alteration of the shape or globe of the eye) From low vision to blindness

slide-11
SLIDE 11

Visual impairment examples

4

Glaucoma (ganglion cells and the optic nerve degenerate) Retinitis Pigmentosa (rods photoreceptors degenerate) Age Related Macular Degeneration (macula cones degenerate) Degenerative myopia (major alteration of the shape or globe of the eye) From low vision to blindness

slide-12
SLIDE 12

Low vision

5

Electronic magnifiers Immersive systems

Main existing tools for accessibility: Magnifiers

Optical devices

slide-13
SLIDE 13

Blindness

6

Existing emerging therapeutic methods

slide-14
SLIDE 14

Blindness

6

Gene therapy Stem cells Cell transplantation Retinal prostheses

(electric, optoelectronic, optogenetic)

Existing emerging therapeutic methods

slide-15
SLIDE 15

Processor Receiver Camera

Blindness

6

Gene therapy Stem cells Cell transplantation Retinal prostheses

(electric, optoelectronic, optogenetic)

Existing emerging therapeutic methods

slide-16
SLIDE 16

Processor Receiver Camera

Blindness

6

Gene therapy Stem cells Cell transplantation Retinal prostheses

(electric, optoelectronic, optogenetic)

Existing emerging therapeutic methods

slide-17
SLIDE 17

Processor Receiver Camera

Blindness

6

Gene therapy Stem cells Cell transplantation Retinal prostheses

(electric, optoelectronic, optogenetic)

Existing emerging therapeutic methods

Photoreceptors

slide-18
SLIDE 18

Processor Receiver Camera

Blindness

6

Gene therapy Stem cells Cell transplantation Retinal prostheses

(electric, optoelectronic, optogenetic)

Existing emerging therapeutic methods

Photoreceptors Ganglion cells

slide-19
SLIDE 19

High scientific, technologic, societal, economic impact New institutes, start-ups, companies However, fundamental issues remain unresolved

7

Helping visually impaired people

Biovision team

slide-20
SLIDE 20

High scientific, technologic, societal, economic impact New institutes, start-ups, companies However, fundamental issues remain unresolved

7

Helping visually impaired people

Biovision team

slide-21
SLIDE 21

Neuroscientists and physicians High scientific, technologic, societal, economic impact New institutes, start-ups, companies However, fundamental issues remain unresolved

7

Helping visually impaired people

Biovision team

Biophysical modelling Mathematical analysis Computer vision Computational neuroscience

slide-22
SLIDE 22

8

Fundamental research

slide-23
SLIDE 23

8

Scene analysis Scene enhancement

(dependent on pathology)

Fundamental research

slide-24
SLIDE 24

8

Methods from computer vision and graphics Scene analysis Scene enhancement

(dependent on pathology)

Fundamental research

slide-25
SLIDE 25

8

Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

Fundamental research

slide-26
SLIDE 26

8

Biophysical modeling Information compression Retina emulation Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

Fundamental research

slide-27
SLIDE 27

8

Biophysical modeling Information compression Retina emulation Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

Feedback Perception

Fundamental research

slide-28
SLIDE 28

Software development

9

slide-29
SLIDE 29

Software development

9

Producing software for biologists and physicians

slide-30
SLIDE 30

Software development

9

Producing software for biologists and physicians

ADT, ANR KEOPS, FP7 Renvision

ENAS: A software for analysing spike trains at single cell and population levels

Cessac et al. (in preparation)

https://enas.inria.fr

Virtual Retina: A large scale simulator of biological retina

Wohrer, Kornprobst (2007)

http://www-sop.inria.fr/neuromathcomp/public/software/virtualretina/

slide-31
SLIDE 31

Software development

9

Producing software for biologists and physicians

ADT, ANR KEOPS, FP7 Renvision

ENAS: A software for analysing spike trains at single cell and population levels

Cessac et al. (in preparation)

https://enas.inria.fr

Virtual Retina: A large scale simulator of biological retina

Wohrer, Kornprobst (2007)

http://www-sop.inria.fr/neuromathcomp/public/software/virtualretina/

slide-32
SLIDE 32

Software development

9

Producing software for biologists and physicians New vision-aid systems for patients with impaired vision

ADT, ANR KEOPS, FP7 Renvision

ENAS: A software for analysing spike trains at single cell and population levels

Cessac et al. (in preparation)

https://enas.inria.fr

Virtual Retina: A large scale simulator of biological retina

Wohrer, Kornprobst (2007)

http://www-sop.inria.fr/neuromathcomp/public/software/virtualretina/

slide-33
SLIDE 33

10

slide-34
SLIDE 34

10

University of Newcastle University of Edinburgh

Receptive fields estimation

RENVISION EU project (ends 2016)

slide-35
SLIDE 35

10

University of Newcastle University of Edinburgh

Receptive fields estimation

RENVISION EU project (ends 2016)

Local edge? Anything moving?

slide-36
SLIDE 36

11

New variational approaches for receptive estimation (ongoing) New method for stimuli design (started) Enas implementation

  • A. Drogoul

Receptive fields estimation

slide-37
SLIDE 37

12

ADT (2012-2014) ANR KEOPS (2011-2015) RENVISION project (ends 2016)

Spike coding

University of Valparaiso Institut de la vision

slide-38
SLIDE 38

12

ADT (2012-2014) ANR KEOPS (2011-2015) RENVISION project (ends 2016)

Spike coding

University of Valparaiso Institut de la vision

Time (ms) Neurons

slide-39
SLIDE 39

13

Spike coding

slide-40
SLIDE 40

13

How do stimuli and connectivity shape the collective retina response?

Spike coding

slide-41
SLIDE 41

13

How do stimuli and connectivity shape the collective retina response? Handling spatio-temporal correlations and non-stationary response to stimuli

Spike coding

slide-42
SLIDE 42

13

How do stimuli and connectivity shape the collective retina response? Handling spatio-temporal correlations and non-stationary response to stimuli Experimental studies and analysis

Spike coding

slide-43
SLIDE 43

13

How do stimuli and connectivity shape the collective retina response? Handling spatio-temporal correlations and non-stationary response to stimuli Experimental studies and analysis Enas implementation

Spike coding

slide-44
SLIDE 44

14

Application to retina prosthesis?

slide-45
SLIDE 45

14

Application to retina prosthesis?

slide-46
SLIDE 46

14

Application to retina prosthesis?

64 pixels - Argus II

Adapted from Institut de la Vision

slide-47
SLIDE 47

14

Application to retina prosthesis?

64 pixels - Argus II

Adapted from Institut de la Vision

256 pixels

slide-48
SLIDE 48

14

Application to retina prosthesis?

64 pixels - Argus II

Adapted from Institut de la Vision

256 pixels 1024 pixels

slide-49
SLIDE 49

15

Application to retina prosthesis?

slide-50
SLIDE 50

15

Application to retina prosthesis?

slide-51
SLIDE 51

16

Scene transforms in computer vision and graphics

Designed for artistic purposes

slide-52
SLIDE 52

16

Scene transforms in computer vision and graphics

Example: Image and video quality improvements: Equalisation, gamma correction, tone mapping, edge enhancement, image decomposition, cartoonization

Winnemoller et al. (2012)

Source XDoG XDoG Thresh.

Designed for artistic purposes

slide-53
SLIDE 53

16

Scene transforms in computer vision and graphics

Example: Image and video quality improvements: Equalisation, gamma correction, tone mapping, edge enhancement, image decomposition, cartoonization

Winnemoller et al. (2012)

Source XDoG XDoG Thresh.

Designed for artistic purposes For retina prosthesis and low vision

slide-54
SLIDE 54

17

Scene transforms in computer vision and graphics

For retina prosthesis and low vision

Example of problem Disambiguate what comes from structure and what comes from illumination

slide-55
SLIDE 55

17

Scene transforms in computer vision and graphics

For retina prosthesis and low vision

Example of problem Disambiguate what comes from structure and what comes from illumination

slide-56
SLIDE 56

18

Biophysical modeling Information compression Retina emulation Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

slide-57
SLIDE 57

19

Biophysical modeling Information compression Retina emulation Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

slide-58
SLIDE 58

19

Biophysical modeling Information compression Retina emulation Methods from computer vision and graphics Retina processing Encoding Stimulation Scene analysis Scene enhancement

(dependent on pathology)

Feedback Perception

slide-59
SLIDE 59

20

ANR Trajectory (M2+PhD, 2015-2018)

  • K. Medathati PhD (2013-2016)

Synergistic model of motion processing

INT Institut de la Vision University of Valparaiso ULM University

slide-60
SLIDE 60

20

ANR Trajectory (M2+PhD, 2015-2018)

  • K. Medathati PhD (2013-2016)

Synergistic model of motion processing

INT Institut de la Vision University of Valparaiso ULM University

Scaling up models rooted in experimental biology (neurophysiology, psychophysics, etc.) leading to an exciting synergy between studies in computer vision and biological vision.

(a)

Retina V1 MT V2 MST V4 LGN

Local m and surfa

Dorsal stream Motion processing Ventral stream Form processing

slide-61
SLIDE 61

21

Fundamental research

Biovision team

Advances in our understanding of the visual system New methods for low vision people New paradigms in computer vision

Multi-scale biophysical modelling Theoretical tools for models analysis Simulations tools New paradigms in vision

slide-62
SLIDE 62

22

Transfert

Biovision team

Retina simulator platform reproducing the collective retinal response for normal and impaired retinas Vision-aid systems for patients with impaired vision Therapeutic strategies for blind patients

slide-63
SLIDE 63

Thank you

https://team.inria.fr/biovision