人工智能深度學(xué)習(xí)課程_第1頁
人工智能深度學(xué)習(xí)課程_第2頁
人工智能深度學(xué)習(xí)課程_第3頁
人工智能深度學(xué)習(xí)課程_第4頁
人工智能深度學(xué)習(xí)課程_第5頁
已閱讀5頁,還剩69頁未讀 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡介

AI

Deep

Learning

ClassFrom

Mystery

to

Masterydef

sigmoid(z):

return

1.0/(1.0+np.exp(-z))Given:

Training

DataTask:

Learn

a

classification

functionLoss

OptimizationWe

v,ant

to

find

the

netwo水v1eigh飛小or

achievethe

lowest

loss1,

nw

=argmin-2L(f(x(i);w),y(i))w

n

i=1w,=

argmin](W)wRemember:w

=

{w(0),,w(1),···

}Loss

Optimizationw·

:=argmin](W)wRemember:Our

loss

is

a

function

ofthe

network

weigh囚丿(w。,W1)。。/飛.掌。W。W1。Loss

OptimizationTake

smalll

step

1n

o,pposite

d1irection

of

gradient_+J(w。,W1)。...I..。W八W,1Gradient

DescentRepeat

until

convergence`3一2-J(w。,W1)。一.2一03o.e

o,-

o;?·飛勹產(chǎn)~,W。°02W1Classification

ProblemHeuristic

Training/Test

Ratio

80/20Heuristic

Pareto

80/20Training

vs

Test

ErrorsNew

Model

with

more

input

dataNew

Model

more

NeuronsNew

Model

with

Noisy

DataImage

creating

a

Face

for

Siri

or

Alexa4/10/2369AI

Contributions

in

the

field

of

AeronauticsOverview4/10/2370Mission

Objectives:

Explore

use

of

asUAS

capability

at

an

aircraft

crashscene

to:Assist

Crash

and

Rescue

EffortsDefine

Crash

Area

–security

and

publicsafetyAssess

hazards

before

entryLocate

personnelVideo

document

the

siteIdentify/locate

objectsGPS

tag/inventory

individual

objectsEnable

management

and

EOC

effortsDetect/Classify

the

Black

BoxStream

video

to

AFRC

Edwardslocation20

MegapixelCameraArtificial

IntelligenceDeep

Neural

NetworkArchitectures4/10/237120

MegapixelCameraGoal:

Object

Detection

and

Classification

BoxBlackTraining

Data:

Over

120

Black

Box

images

(2D)

on

the

internetCNN

Model:

Yolo

detection

network

has

24

convolutional

layers

deep

followed

by

2fully

connectedlayers.----Black

Box----

Aircrafts----

PeopleAI

CNN

Detection

Accuracy72

Results:

YOLOv2predicts

(96%)

bounding

boxon

1080P

resolution

at

30FPSvideoat

low

altitudesNeural

Networks

for

Object

Detection

and

Tracking

(speed,

accuracy,

and

efficiency)Aircraft

Crash

MishapExercise4/10/2373Results:

Video

Uploaded

2.5

Mbps

at

30

FPS

~

3

second

lag

on

Live

Stream

Video

on

MiFiMission

Economics

based

on

the

hazardous

conditions

and

limited

time4.Accurac

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

最新文檔

評論

0/150

提交評論