Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Why so serious? Using GCP for realtime video co...

Avatar for Johann du Toit Johann du Toit
March 06, 2019
Programming
89
0

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

Avatar for Johann du Toit

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
Avatar for Johann du Toit johanndutoit
0
29
RESPECTING THE MICROSERVICE.
Avatar for Johann du Toit johanndutoit
1
87
Attacking a K8s cluster and how defending it
Avatar for Johann du Toit johanndutoit
0
73
Progressive Web Apps in 15 Minutes
Avatar for Johann du Toit johanndutoit
0
52
The Machine Learning Intervention
Avatar for Johann du Toit johanndutoit
0
75
Progressive Web Apps
Avatar for Johann du Toit johanndutoit
0
100
Progressive Web Apps
Avatar for Johann du Toit johanndutoit
0
110
Launchpad Start Day 2 - Google Design Sprints
Avatar for Johann du Toit johanndutoit
2
140
Google Design Sprints
Avatar for Johann du Toit johanndutoit
0
99

Other Decks in Programming

See All in Programming
Migrations : C'est une question d'hygiène !
Avatar for Vincent Amstoutz vinceamstoutz
0
1.8k
OCRを使ってゲームのアイテムをデータ化する
Avatar for Kishikawa Katsumi kishikawakatsumi
0
120
プロパティの順序で型推論が壊れる!? TypeScript6.0の修正からContext-Sensitivityの仕組みを追う
Avatar for おおいし bicstone
2
1k
Modding RubyKaigi for Myself
Avatar for yui-knk yui_knk
0
440
開発体験を左右するライブラリの API 設計 - GraphQL スキーマ構築ライブラリから考える #tskaigi
Avatar for izumin5210 izumin5210
2
530
inferと仲良くなる10分間
Avatar for ryokatsuse ryokatsuse
1
250
LLM Plugin for Node-REDの利用方法と開発について
Avatar for 404background 404background
0
110
Hive Metastoreを通して学ぶIceberg REST Catalog ― 仕様から実装まで
Avatar for okumin okumin
0
280
TypeScriptだけでAIエージェントを作る フロント・エージェント・インフラのフルスタック実践
Avatar for Har1101 har1101
6
1k
Zod v4 Codec でスキーマに型変換を埋め込む REST API 設計 #TSKaigi2026
Avatar for Ryutaro Yako ryutaro_yako
0
150
RailsTokyo 2026#4: AI様があれば、
Hotwireの弱点は消えるか？
Avatar for Naofumi Kagami naofumi
5
980
柔軟なPDFレイアウトエディタを支える型システム設計 — Discriminated UnionとConditional Typeの実践
Avatar for minako-ph minako__ph
4
590

Featured

See All Featured
Chasing Engaging Ingredients in Design
Avatar for Sebastian Deterding codingconduct
0
200
Primal Persuasion: How to Engage the Brain for Learning That Lasts
Avatar for Mike Taylor tmiket
0
340
A Modern Web Designer's Workflow
Avatar for Chris Coyier chriscoyier
698
190k
Agile Actions for Facilitating Distributed Teams - ADO2019
Avatar for Mark Kilby mkilby
0
190
Navigating Algorithm Shifts & AI Overviews - #SMXNext
Avatar for Aleyda Solis aleyda
1
1.2k
The Mindset for Success: Future Career Progression
Avatar for Greg Gifford greggifford
PRO
0
340
Hiding What from Whom? A Critical Review of the History of Programming languages for Music
Avatar for Tomoya Matsuura tomoyanonymous
2
820
The Cult of Friendly URLs
Avatar for Andy Hume andyhume
79
6.9k
Fight the Zombie Pattern Library - RWD Summit 2016
Avatar for Marcelo Somers marcelosomers
234
17k
DBのスキルで生き残る技術 - AI時代におけるテーブル設計の勘所
Avatar for soudai sone soudai
PRO
65
54k
Refactoring Trust on Your Teams (GOTO; Chicago 2020)
Avatar for Rebecca Miller-Webster rmw
35
3.5k
Site-Speed That Sticks
Avatar for Harry Roberts csswizardry
13
1.2k

Transcript

  1. None

  2. [email protected] email website @signedness

  3. Technical Lead for Cloud

  4. Developer Expert (first in Africa) https://developers.google.com/experts/people/johann-du-toit

  5. Certified Architect + Data Engineer

  6. Mentor since 2016

  7. None

  8. THIS TALK Realtime processing on video Further processing using Vision

    API Training a custom model on Google Cloud

  9. THIS TALK Realtime processing on video Further processing using Vision

    API Training a custom model on Google Cloud
  10. None

  11. 14 cameras per metro station on your way out

  12. None
  13. None
  14. None
  15. None
  16. None
  17. None

  18. gocv.VideoCaptureDevice(0)

  19. gocv.VideoCaptureFile(“a.mp4”)

  20. gocv.IMRead(“test.png”)

  21. for { webcam.Read(&img) window.IMShow(img) window.WaitKey(1) }

  22. for { webcam.Read(&img) window.IMShow(img) window.WaitKey(1) }

  23. gocv.NewWindow("Raw Video Source")

  24. gocv.NewWindow("Raw Video Source") gocv.NewWindow("Raw Video Source")

  25. None

  26. for { webcam.Read(&img) // identify cars window.IMShow(img) window.WaitKey(1) }

  27. None
  28. None
  29. None

  30. 30 frames per second

  31. 1000ms

  32. 33ms

  33. None

  34. ~20ms

  35. None
  36. None
  37. None

  38. 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
  39. None
  40. None
  41. None
  42. None

  43. HAAR Cascades

  44. HAAR Wavelets

  45. None
  46. None
  47. None
  48. None

  49. ~20ms

  50. None

  51. frameNumber % 1 == 0

  52. frameNumber % 2 == 0

  53. frameNumber % 3 == 0

  54. None

  55. THE AGENDA Realtime processing on video Further processing using Vision

    API Training a custom model on Google Cloud
  56. None
  57. None
  58. None

  59. 2 Events

  60. None
  61. None

  62. ImageNet (14 million images)

  63. > day on

  64. < 6 minutes

  65. None
  66. None
  67. None
  68. None
  69. None
  70. None
  71. None
  72. None
  73. None
  74. None
  75. None
  76. None
  77. None
  78. None
  79. None
  80. None

  81. THE AGENDA Realtime processing on video Vision API Training a

    custom model on Google Cloud
  82. None
  83. None
  84. None
  85. None
  86. None
  87. None
  88. None
  89. None
  90. None
  91. None
  92. None
  93. None
  94. None
  95. None
  96. None
  97. None

  98. FIN

  99. FINN