Categorize image regions

[flowac]

1. Detect / select interviewee and text regions

Note: Probably easiest to select regions / windows for processing. Unfortunately OS dependent.
Note: Might be more appropriate for Phase 3

[flowac]

Standalone natively ran program (select window to record).
Python or C++ taking periodic snapshots. Either process locally or send to server.

Vendor integration could be helpful but too many possible configurations.

[flowac]

Currently using QT6 to display window list. It should be able to hijack audio stream too.

[flowac]

QT6 GDI capture tool archived. Windows 10 based DX11 tool added.

[flowac]

Image capture nearly complete. Currently having file path issues. Audio capture will be handled in a separate issue.

[flowac]

Client side implemented in Commit 5119fcf
Took hours to do a simple file open using winrt