~/talk

Building computer vision applications Q&A

on YouTube 46m
Talkat Microsoft Build 2023 · 2023-05-23..05-25· [computer-vision, machine-learning]

speakers

name:Kellen Anderson
links:none on file
name:Courtney Brewer
links:none on file
name:Kate Browne
links:none on file
name:Shawndell Elfring
linkedin:shawndell-elfring-b32349b4
name:Tonye Iyalla
links:none on file
name:Fisayo Okikiolu
links:none on file
name:Cenyu Zhang
links:none on file

resources

video:https://youtu.be/9kUcOiDb2DU
slides:none on file

summary

Azure Cognitive Services for Vision offers captioning (full image or up to 10 regions), vector-based retrieval and video search, background removal/segmentation, OCR/Read API including handwriting, and a new video summarization that already summarizes sample uploads; customization uses the Florence foundation model for few‑shot product and object recognition, face verification uses Recognition_04 face model, body‑tracking (32 joint points) and edge containers are in private preview, Vision Studio allows testing but not model export, and customer training data stays private.

AI-written summary · about

chapters (28)

  1. 0:00Q&A
  2. 1:17Azure Cognitive Services for Vision
  3. 5:19OCR and real-time vision
  4. 6:49CNN alternatives for classification
  5. 8:20OCR text and file size limits
  6. 9:04ID verification and facial recognition
  7. 11:33Training object detection models
  8. 13:02Aging algorithms for missing children
  9. 14:48Video summarization feedback
  10. 16:15Spectral imaging and custom models
  11. 18:48Product type and serial number detection
  12. 20:40Vision analytics for equipment failure
  13. 23:10Importing and exporting models
  14. 23:52Body tracking and spatial analysis
  15. 25:40Florence vs GPT-4 image comparison
  16. 27:27Multimodal vision, language, and sound
  17. 28:31Ingesting video feeds
  18. 31:05Proprietary data and model policies
  19. 33:12Azure Depth Platform and 3D services
  20. 34:16HoloLens hand tracking
  21. 35:21Low frame rate image simulation
  22. 37:35Edge support for custom models
  23. 38:57Edge support timeline
  24. 40:02Passive identification in care facilities
  25. 41:52Video transcription and script generation
  26. 43:22Video vs image search differences
  27. 44:10Spatial analysis with sequential images
  28. 45:57Closing remarks

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.