8.0 object recognition - cs.cmu.edu16385/s17/slides/8.0_object_recognition.pdf · object...
TRANSCRIPT
![Page 1: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/1.jpg)
Object Recognition16-385 Computer Vision (Kris Kitani)
Carnegie Mellon University
Henderson and Davis. Shape recognition using hierarchical
Constraint Analysis. 1979
![Page 2: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/2.jpg)
What do we mean by ‘object recognition’?
![Page 3: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/3.jpg)
Is this a street light?(Verification / classification)
![Page 4: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/4.jpg)
Where are the people?(Detection)
![Page 5: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/5.jpg)
Is that Potala palace?(Identification)
![Page 6: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/6.jpg)
What’s in the scene?(semantic segmentation)
Building
Mountain
Trees
VendorsPeople
Ground
Sky
![Page 7: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/7.jpg)
What type of scene is it?(Scene categorization)
Outdoor
City
Marketplace
![Page 8: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/8.jpg)
Challenges (Object Recognition)
![Page 9: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/9.jpg)
Viewpoint variation
![Page 10: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/10.jpg)
Illumination variation
![Page 11: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/11.jpg)
Scale variation
![Page 12: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/12.jpg)
Background clutter
![Page 13: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/13.jpg)
Deformation
![Page 14: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/14.jpg)
Occlusion
![Page 15: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/15.jpg)
Intra-class variation
![Page 16: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/16.jpg)
Common approaches
![Page 17: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/17.jpg)
Spatial reasoning
Window classification
Feature Matching
Common approaches: object recognition
![Page 18: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/18.jpg)
Feature matching
![Page 19: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/19.jpg)
What object do these parts belong to?
![Page 20: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/20.jpg)
a collection of local features (bag-of-features)
An object as
Some local feature are very informative
Are the positions of the parts important?
• deals well with occlusion • scale invariant • rotation invariant
![Page 21: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/21.jpg)
Pros
• Simple
• Efficient algorithms
• Robust to deformations
Cons
• No spatial reasoning
![Page 22: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/22.jpg)
Spatial reasoning
Window classification
Feature Matching
Common approaches: object recognition
![Page 23: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/23.jpg)
Spatial reasoning
![Page 24: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/24.jpg)
The position of every part depends on the positions of all the other parts
positional dependence
Many parts, many dependencies!
![Page 25: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/25.jpg)
1. Extract features 2. Match features 3. Spatial verification
![Page 26: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/26.jpg)
1. Extract features 2. Match features 3. Spatial verification
![Page 27: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/27.jpg)
1. Extract features 2. Match features 3. Spatial verification
an old idea…
![Page 28: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/28.jpg)
Fu and Booth. Grammatical Inference. 1975
Structural (grammatical) descriptionScene
![Page 29: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/29.jpg)
![Page 30: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/30.jpg)
1972
Description for left edge of face
![Page 31: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/31.jpg)
vector of RVs: set of part locations L = {L1, L2, . . . , LM}
RV
A more modern probabilistic approach…
think of locations as random variables (RV)
RV RV
![Page 32: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/32.jpg)
vector of RVs: set of part locations L = {L1, L2, . . . , LM}
What are the dimensions of R.V. L?
How many possible combinations of part locations?
RV
A more modern probabilistic approach…
think of locations as random variables (RV)
RV RV
L1
L2
LM
image (N pixels)
![Page 33: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/33.jpg)
vector of RVs: set of part locations L = {L1, L2, . . . , LM}
What are the dimensions of R.V. L?
How many possible combinations of part locations?
RV
A more modern probabilistic approach…
think of locations as random variables (RV)
RV RV
L1
L2
LM
image
Lm = [ x y ]
![Page 34: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/34.jpg)
vector of RVs: set of part locations L = {L1, L2, . . . , LM}
What are the dimensions of R.V. L?
How many possible combinations of part locations?
NM
RV
A more modern probabilistic approach…
think of locations as random variables (RV)
RV RV
Lm = [ x y ]
L1
L2
LM
image
![Page 35: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/35.jpg)
Most likely set of locations L is found by maximizing:
Likelihood: How likely it is to observe
image I given that the M parts are at locations L
(scaled output of a classifier)
Prior: spatial prior controls the
geometric configuration of the parts
What kind of prior can we formulate?
p(L|I) / p(I|L)p(L)part
locations image
Posterior
![Page 36: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/36.jpg)
Given any collection of selfie images, where would you expect the nose to be?
P (Lnose
) =?
What would be an appropriate prior?
![Page 37: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/37.jpg)
A simple factorized model
Break up the joint probability into smaller (independent) terms
p(L) =Y
m
p(Lm)
![Page 38: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/38.jpg)
Independent locations
Each feature is allowed to move independently
Does not model the relative location of parts at all
p(L) =Y
m
p(Lm)
![Page 39: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/39.jpg)
Tree structure (star model)
Represent the location of all the parts relative to a single
reference part
Assumes that one reference part is defined
(who will decide this?)
Root (reference)
node
p(L) = p(Lroot
)M�1Y
m=1
p(Lm|Lroot
)
![Page 40: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/40.jpg)
Fully connected (constellation model)
Explicitly represents the joint distribution of locations
Good model: Models relative location of parts
BUT Intractable for moderate number of parts
positional dependence
p(L) = p(l1, . . . , lN )
![Page 41: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/41.jpg)
Pros
• Retains spatial constraints
• Robust to deformations
Cons
• Computationally expensive
• Generalization to large inter-class variation (e.g., modeling chairs)
![Page 42: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/42.jpg)
Spatial reasoning
Window classification
Feature Matching
![Page 43: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/43.jpg)
Window-based
![Page 44: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/44.jpg)
1. get image window 2. extract features 3. classify
When does this work and when does it fail?
How many templates do you need?
Template Matching
![Page 45: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/45.jpg)
Per-exemplar
find the ‘nearest’ exemplar, inherit its label
exemplar template top hits from test data
![Page 46: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/46.jpg)
1. get image window(or region proposals)
2. extract features 3. compare to template
Template Matching
Do this part with one big classifier ‘end to end learning’
![Page 47: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/47.jpg)
Convolutional Neural Networks
Convolution Pooling
Image patch (raw pixels values)
response of one ‘filter’
max/min response over
a region
A 96 x 96 image convolved with 400 filters (features) of size 8 x 8 generates about 3
million values (892x400)
Pooling aggregates statistics and lowers the dimension of convolution
Image patch (raw pixels values)
response of one ‘filter’
![Page 48: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/48.jpg)
![Page 49: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/49.jpg)
630 million connections 60 millions parameters to learn
Krizhevsky, A., Sutskever, I. and Hinton, G. E. ImageNet Classification with Deep Convolutional Neural Networks, NIPS 2012.
224/4=56
96 ‘filters’
![Page 50: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/50.jpg)
Pros
• Retains spatial constraints
• Efficient test time performance
Cons
• Many many possible windows to evaluate
• Requires large amounts of data
• Sometimes (very) slow to train
![Page 51: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/51.jpg)
How to write an effective CV resume
![Page 52: 8.0 Object Recognition - cs.cmu.edu16385/s17/Slides/8.0_Object_Recognition.pdf · Object Recognition 16-385 Computer Vision (Kris Kitani) Carnegie Mellon University Henderson and](https://reader033.vdocuments.mx/reader033/viewer/2022042912/5f488620e2d5da0979339052/html5/thumbnails/52.jpg)
Deep Learning+1-DEEP-LEARNING deeplearning@deeplearning http://deeplearning
Summary: Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning
Experience: Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Education
Deep Learning Deep Learning ?Deep Learning Deep Learning Deep Learning
Experience
Deep Learning Deep Learning .Deep Learning Deep Learning, Deep Learning
· Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
· Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep
Learning Deep Learning Deep Learning
· Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep
Learning Deep Learning
Deep Learning in another country Deep LearningDeep Learning , Deep Learning , Deep Learning
· Deep Learning ... wait.. Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning
Deep Learning Deep Learning Deep Learning
· Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep Learning Deep
Learning Deep Learning
Deep Learning Deep LearningDeep Learning Deep Learning
· Very Deep Learning
Publications
1. Deep Learning in Deep Learning People who do Deep Learning things. Conference of Deep
Learning.
2. Shallow Learning... Nawww.. Deep Learning bruh Under submission while Deep Learning
Patent
1. System and Method for Deep Learning. Deep Learning, Deep Learning , Deep Learning , Deep
Learning
1