Fully Convolutional Networks for Image Segmentation | SciPy 2017 | Daniil Pakhomov

[@daniilpakhomov5596]

Slides are available here (read comments for the slides, they have more details):
docs.google.com/presentation/d/1fZnWsrnjNCENZVk2q2ulsvljlamHCsLA5vvWE-SQ3sU/edit?usp=sharing

[@tymothylim6550]

Thank you very much for this video! It was very helpful for me to understand Fully Convolutional Networks!

[@SAINIVEDH]

Great talk. Where can I learn more about upsampling methods.

[@enthought]

See the complete SciPy 2017 Conference talk & tutorial playlist here: ruclips.net/p/PLYx7XA2nY5GfdAFycPLBdUDOUtdQIVoMf

[@anmolagar]

How do i create ground truth? Because i want to work on completely new set of classes. What is the best approach?

[@paulmaksud3330]

it's been two months so i guess you have the answer but you can use software like vgg image annotator to annotate images and produce coordinate of mask.

[@daniilpakhomov5596]

There is one way to do this using this:
github.com/warmspringwinds/dash-deep

[@myintmyintmaw2621]

Thank you for your explanations.

[@lartsevevgenii6640]

Good lesson thank you so much!

[@amessit10]

what will be the approximate cost if we go for a project at engineering level ?

[@10monthspregnant]

$100

[@amessit10]

momonala what will be hardware if I wish to do project as home security system?

[@dho]

xbox

[@karrikarthik6936]

Are there any written notes for this presentation?

[@daniilpakhomov5596]

docs.google.com/presentation/d/13q5IZRJHs-Rdeag_AwINpGmcF8aCCr-Vp_KZvmVkp8k/edit?usp=sharing
The presentation has notes

[@vinithapalani3385]

@@daniilpakhomov5596 Thank you very much

[@КириллПинигин-в3е]

Даниил респект тебе:)

[@trumpthetraitor2485]

Great Video.