Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Why so serious? Using GCP for realtime video co...
Search
Johann du Toit
March 06, 2019
Programming
90
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
Why so serious? Using GCP for realtime video context analyzation
Look into real-time video processing and how you might do this on Google Cloud
Johann du Toit
March 06, 2019
More Decks by Johann du Toit
See All by Johann du Toit
Natural Language, Unnatural Access: The Emerging LLM Attack Pattern
johanndutoit
0
37
RESPECTING THE MICROSERVICE.
johanndutoit
1
91
Attacking a K8s cluster and how defending it
johanndutoit
0
78
Progressive Web Apps in 15 Minutes
johanndutoit
0
65
The Machine Learning Intervention
johanndutoit
0
84
Progressive Web Apps
johanndutoit
0
110
Progressive Web Apps
johanndutoit
0
120
Launchpad Start Day 2 - Google Design Sprints
johanndutoit
2
140
Google Design Sprints
johanndutoit
0
100
Other Decks in Programming
See All in Programming
SlackアプリとLambdaの 連携を構築した話
pawn_4_s
1
120
Japan Community Day at Kubecon + CloudNativeCon Japan 2026: Learning Container Privilege Control by Building My Own Low-Level Container Runtime
ternbusty
1
160
ソフトウェア設計に溶けるインフラ ― AWS CDK のインフラ認識論
konokenj
3
800
Laravelで学ぶ Webアプリケーションチューニング入門/web_application_tuning_101
hanhan1978
4
1.9k
楽しそうなつよつよエンジニアと目が死んでる僕/A brilliant engineer having a blast, and dead-eyed me.
3l4l5
2
190
源内ハンズオン概要編
hideg
0
190
改善しないと、タスクが回らない。 “てんこ盛りポジション” を引き継いだ情シスの、入社3ヶ月の業務改善録
krm963
0
270
KotlinConf Extended South Korea 2026 Keynote
l2hyunwoo
0
110
ソフトウェアエンジニアにとっての生成AI - 特性を知って使い倒す / generative ai for software enginner
kishida
5
1.7k
20260722_microCMSで考える、AI時代のコンテンツ運用設計
yosh1
0
410
「人を評価する AI」の設計と実装
ryoyanara
0
210
What's New in Android 2026
veronikapj
0
260
Featured
See All Featured
Designing for humans not robots
tammielis
254
26k
Future Trends and Review - Lecture 12 - Web Technologies (1019888BNR)
signer
PRO
0
3.7k
Leveraging Curiosity to Care for An Aging Population
cassininazir
1
470
Winning Ecommerce Organic Search in an AI Era - #searchnstuff2025
aleyda
1
2.1k
A Guide to Academic Writing Using Generative AI - A Workshop
ks91
PRO
1
370
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
12
1.8k
The World Runs on Bad Software
bkeepers
PRO
72
12k
Organizational Design Perspectives: An Ontology of Organizational Design Elements
kimpetersen
PRO
1
790
職位にかかわらず全員がリーダーシップを発揮するチーム作り / Building a team where everyone can demonstrate leadership regardless of position
madoxten
64
56k
Design in an AI World
tapps
1
280
HDC tutorial
michielstock
2
790
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
49
10k
Transcript
None
[email protected]
email website @signedness
Technical Lead for Cloud
Developer Expert (first in Africa) https://developers.google.com/experts/people/johann-du-toit
Certified Architect + Data Engineer
Mentor since 2016
None
THIS TALK Realtime processing on video Further processing using Vision
API Training a custom model on Google Cloud
THIS TALK Realtime processing on video Further processing using Vision
API Training a custom model on Google Cloud
None
14 cameras per metro station on your way out
None
None
None
None
None
None
gocv.VideoCaptureDevice(0)
gocv.VideoCaptureFile(“a.mp4”)
gocv.IMRead(“test.png”)
for { webcam.Read(&img) window.IMShow(img) window.WaitKey(1) }
for { webcam.Read(&img) window.IMShow(img) window.WaitKey(1) }
gocv.NewWindow("Raw Video Source")
gocv.NewWindow("Raw Video Source") gocv.NewWindow("Raw Video Source")
None
for { webcam.Read(&img) // identify cars window.IMShow(img) window.WaitKey(1) }
None
None
None
30 frames per second
1000ms
33ms
None
~20ms
None
None
None
10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms
10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10m 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10m 10ms 10ms 10m 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10ms 10m 10m 10ms 10ms 10ms 10ms
None
None
None
None
HAAR Cascades
HAAR Wavelets
None
None
None
None
~20ms
None
frameNumber % 1 == 0
frameNumber % 2 == 0
frameNumber % 3 == 0
None
THE AGENDA Realtime processing on video Further processing using Vision
API Training a custom model on Google Cloud
None
None
None
2 Events
None
None
ImageNet (14 million images)
> day on
< 6 minutes
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
THE AGENDA Realtime processing on video Vision API Training a
custom model on Google Cloud
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
FIN
FINN