Thanks to visit codestin.com
Credit goes to www.scribd.com

0% found this document useful (0 votes)
25 views7 pages

CNN Notes

The document discusses Convolutional Neural Networks (CNNs) and their architecture, focusing on how they process input data through convolutional layers, activation functions, and pooling layers. It explains the role of kernels in feature extraction and the use of ReLU for non-linearity. Additionally, the document touches on the final classification step using softmax to determine output probabilities.

Uploaded by

akashsingh38671
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
25 views7 pages

CNN Notes

The document discusses Convolutional Neural Networks (CNNs) and their architecture, focusing on how they process input data through convolutional layers, activation functions, and pooling layers. It explains the role of kernels in feature extraction and the use of ReLU for non-linearity. Additionally, the document touches on the final classification step using softmax to determine output probabilities.

Uploaded by

akashsingh38671
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Conyolulional Neual Netwoulk (aNN)

that
s a

a t pattins

tex tuucs to
yowr bau docnr
lek a a pie. of a d
laxluoa,
pirel. falead

CNN Aveiteelia
convolulona!

( deuae
@onneeid

Convolulona!
fooliig
dhput Laye lao tayt

How
Llo the
The
tales au ble.
) The eo
a lue
Vale uot ike
) Each pixel enen the
Augle
ooloumd kou

Lalli ees

A 4x4 Aeale

H
10 30 40
50 60 70 80
40 |20
(30 (60

Black
Pirel Vauu
Pixel Valu 5’

Same fos
Peduni (Convelelea
i) Jhe NN wes iy
to Aean the
uage aud delu

baie

at
tuough a awal
teokng time. lhale sat

'does

a dua! 04 Keunel that ides ovey


CNN
7he
a aloo a
the iruage. he filli bo uage aud he ffrla.
a dot puadu et
Eg 8 Lete day e
eonvolulebn te
LGu hes to
bo plaeed
ne Vaue. Move to 291L nQusho add theu to
next p0AI tton au
fos the
top-leyt 3x3
w e R e a t ,

50 0 30
70
|stecole
|0D meving fA

he neeo Vollue

30
50
-

Ts ne0
ne

Sep 30
Removeig Unnscsonauy Aeki (Relu AeÃvalün,
) 4Louading palku , te CNN Aeaoes the unneALAany
detit the peutoud paui
ReLu aufou
) t dou thu ning a tune. ealled OR!.
uQak

to
phen
alloulion to to

Vales
Aleu eonVo lalion AOLe
plau negatie valus lk .
furdlt'n
) max (0,2)
30

20 I5

Aflu

30

Poelin

Puolueca

pfo tuse . Even sheng


au
duale

) Poolig helps to made


ifosualzon
") he uest 0 MLen me thed lar

Ret0 f e get a 4xe ualiâ


guen kalo w
hen divoon oiu be ha.
30
I00 50

30
value

in each.

30

No

Conneeled

A
a Lat fen

leali n te
DO ties to

pfelau.

CNN
kge to cat.
leo fleneo
eonfoeuto valuo
value
to oleg auo
ID vee to.
faliu wap ulo a
) he eN flalis the Conne
Qul eneh

conne es to
) aud odded
Value

90, 90]
X 30,, t0D,
ealeulald as

addad
a bial) uy aloo be
to he laat loyo.

)a funelið n Softnax to deleuniae


in tu iuage. gt

90%, eat -5
) NN uaes like doftna x to doeude He
tune.
elas. Jhe lo

Selthuar alulati+
P(4)

2.5

e2:5412 t31

P(y)
P()
the

CNN padea fu aolog


SEudo

Asagnment 2

tualencat

n <3 leuel

n> 7’ leve L2

Sho

You might also like