Video Tutorials - Face Recognition

Introduction

Below more video tutorials...

Can't be done, doesn't exist - Synchronization between cloned User Accounts

...a unique feature of Streams and Hubzilla.



00:00 Synchronization of cloned user accounts has been running stable for years
00:30 A primary user account and its clone on a different server
01:00 Many clones are possible
01:35 Detected faces and names of faces are synchronized between clones
02:40 Pictures (uploaded) are synchronized
03:40 Permissions are synchronized (example picture)
04:50 Name of face is synchronized (again)
05:20 Posts are synchronized
05:45 Send direct message from another account
06:40 The post and the comments are synchronized

Beginners Guide

Touches many topics.



00:00 summary - visual - what will be demonstrated
01:10 summary - recognized faces are synchronized between cloned accounts (server 1 <-> server 2)
01:50 summary - as text
02:30 register an account at server number 1
03:50 upload profile photo
04:40 a look through the settings
04:50 setting - security and privacy
05:30 setting - notifications
05:45 admin - setting - registration without mail verification (fake mail address possible)
06:30 setting - narrow navigation bar
06:35 setting - block channels, block servers
06:45 export channel - to file (as backup or to register again after removal of account)
07:30 addons - find and install addons
07:45 addon Faces - activate addon faces
08:00 addon Faces - how to open and pin it to the navbar
08:20 addon Faces - start for the very first time
08:35 upload pictures - photo app
09:00 upload pictures - post
09:25 upload pictures - post - permission
09:35 upload pictures - files app - permissions
10:00 permission on files - set permission of a single picture ("only me")
10:25 permission on files - demonstrate the effect for an observer - hide a single picture
10:55 permission on files - set permission for directories including files and subfolders ("only me")
11:25 permission on files - demonstrate the effect for an observer - hide all pictures
11:35 addon Faces - start face detection
12:00 addon Faces - name faces
12:10 addon Faces - name faces - use accounts/ contacts that are unique in the decentralized network
12:15 addon Faces - name faces - use any name (just a letters)
12:40 addon Faces - share faces with close contacts only (default: nothing is shared with anybody)
12:50 addon Faces - demonstrate the effect of permissions
13:00 addon Faces - effect of permission - an observer sees no image (set to "only me")
13:15 permission on files - set permission for directories including files and sub folders to "public"
13:30 addon Faces - effect of permission - an observer sees all images again
14:00 clone account - from server 1 to server 2
14:25 clone account - register an account at a server number 2
15:20 admin - setting - registration without mail verification (fake mail address possible)
15:30 clone account - import an existing channel - from server number 1
15:45 clone account - import an existing channel - from file would be possible
15:55 clone account - import an existing channel - use the old (existing) channel address at server number 1
16:10 clone account - completed
16:15 admin - setting - deactivate new registrations
16:35 clone account - login with original account at server number 1
16:50 clone account - demonstrate synchronization between clones (posts, files, permissions, contacts,...)
17:20 addon Faces - activate for clone at server number 2
17:40 addon Faces - demonstrate the synchronization of faces between clones (server 1 <-> server 2
18:25 bonus - posts - format text using markdown (headlines, bold, lists,...)
19:30 bonus - posts - permissions
19:40 bonus - posts - special markdown - replacements of observer url and observer name
20:20 bonus - posts - end-to-end encryption (addon Secrets)
22:00 bonus - find a channel via app "Directory"
22:50 bonus - connect to contacts (to allow comments)
23:44 the end

Share Faces with Friends

You are in full control of how can use your faces.



00:00 Imagine you use the face recognition for the very first time
00:15 Magic - your faces carry names - WOW
00:30 ...because your friends named faces for you
00:35 conditions for sharing faces - a) addon Faces installed, b) you are close friends
00:55 no faces are shared by default
01:05 share faces over server boundaries and even between cloned user account
01:50 share a face from friend to friend - first demo
02:25 only faces named with a contact are shared
02:55 install the app "Friend Zoom
03:40 set (the closeness of) relationship below 50 (0 - 100)
04:50 faces are shared now
05:40 effect of close relationships - contacts are in drop-down list of names
06:20 set back relationship to the default (closeness) of 50
07:00 faces are not shared any longer
07:50 page to check who is sharing what faces
08:05 who is sharing faces with you
08:20 faces you share
08:30 faces others share with you
08:40 moving the affinity slider - the effects
09:20 share faces with contacts that are not close friends
09:40 some technical stuff (what data files are used,...)

WebDAV - Let your Computer do the hard Work

Connect your computer to the cloud files. Run the face detection and recognition on your computer (to reduce server load).



00:00 Introduction and summary
00:30 Introduction and summary - admin - block the detection (python) on the server
00:45 Introduction and summary - run the demo
02:05 This tutorial uses two servers and two clones of a user account
02:40 Admin - first server - deactivate the detection and recognition on the server (block Python scripts)
03:25 Admin - second server - same
03:45 Page - settings - URLs to attach the cloud files as network drive
04:30 Page - help - step-by-step instructions
05:15 webDAV - script - there is one to configure webDAV for Linux
05:40 webDAV - script - the parameters
07:25 webDAV - script - run
07:40 webDAV - script - what it does - file /etc/fstab
08:00 webDAV - script - what it does - create mount directory
08:20 webDAV - script - what it does - file secrete (user credentials)
08:45 webDAV - script - command to connect and disconnect to the cloud files
09:10 webDAV - terminal - connect to cloud files
09:20 webDAV - terminal - cloud files are accessible
09:30 webDAV - file manager - cloud files are accessible
09:55 webDAV - terminal - disconnect cloud files
10:30 webDAV - terminal - connect
10:45 Page - help - install python modules
11:20 Page - help - download and install script for face detection and recognition
11:30 Script - clone repository (download)
13:05 Script - check environment
13:45 Script - help - parameters
13:55 Script - parameter - directory - find the images and config file
14:30 Script - parameter - logfile - print out messages
14:45 Script - parameter - exclude - directories like trash
14:55 Script - parameters - all used
15:20 Page - settings - create config file
16:30 Files involved
17:00 Script - run
17:35 Results - files
17:55 Results - browser - faces are detected / shown - cloned user account
18:00 Results - browser - faces are detected / shown - primary user account
18:05 Addon - face - name a face
18:15 Addon - face - synchronized to the clone
18:35 Script - run
18:45 Addon - face - recognized in other images - primary user account
18:55 Addon - face - recognized in other images - cloned user account
18:05 Addon - face - name a face - real contact
19:25 File - names.json - used to hand over a name to the script
19:40 Script - run
19:55 Addon - face - recognized in other images - primary user account
20:05 Addon - face - recognized in other images - cloned user account
20:15 Files - directories - alongside images
20:45 Files - directories - faces.json - content
21:15 Files - directory faces - share.json - used to share faces with contacts - content
22:35 Files - directory faces - share.json - representation of a face
23:05 Files - directory faces - share.json - representation of a face are compared

The Settings Page

Set detectors, models, minimum size of faces and much more.

(Please connect if you want to see the video.)



00:00 Overview
01:25 Admin - switch on - "experimental" and "immediately"
01:55 Admin - switch on/off- effect for user
02:30 Start tutorial
02:40 How to open the settings page
03:05 Share faces - dedicated page
04:00 Share faces - demo - who is sharing faces depending on affinity slider
04:40 Browser appearance - filter and sort
04:55 Browser appearance - exif - the date when a picture was taken
05:30 Insights - switch on - models and faces
06:10 Insights - faces - show content of file (internal data of face recognition)
07:40 Browser appearance - filter by time - last modified (default)
08:00 Browser appearance - filter by time - exif - when a picture was taken
08:40 Browser appearance - filter by time - exif  - effect of date filter
09:15 Browser appearance - sort by time
10:00 Browser appearance - immediat search (more a behavior of the recognition than appearance)
10:40 Detectors and recognition models - what do they do
11:30 Detectors and recognition models - distance metrics - how faces are found
11:40 Detectors and recognition models - file - face statistics
12:30 Detectors - enable a second detector
13:30 Detectors - enable a second detector - effects - file - face statistics
14:10 Recognition models - enable a second model
15:00 Recognition models - enable a second model - effects - file - face statistics
15:40 Distance metrics - file - faces statistics
16:15 Distance metrics - enforce to use all distance metrics - how and recommendation
17:05 Facial attributes - switch all on
18:00 Facial attributes - switch all on - effects - file - face statistics
18:35 Facial attributes - not used at the moments - files where they are stored
19:35 Remove faces
19:45 Remove faces - everything is stored in files (by the addon)
19:50 Remove faces - page
20:00 Remove faces - page - remove all results of a detector
20:10 Remove faces - page - remove all results of a recognition model
20:25 Remove faces - effects
20:45 Detection - tune
20:50 Detection - tune - minimum size of faces - set
22:00 Detection - tune - minimum size of faces - effects - browser - what the user sees
22:10 Detection - tune - minimum size of faces - effects - file - faces statistics
22:30 Delete data files to remove all results of detection and recognition
23:30 Recognition - tune
24:10 Recognition - tune - training data - minimum size of faces - set - value to big
24:50 Recognition - tune - training data - minimum size of faces - effect - no face recognized
25:35 Recognition - tune - training data - minimum size of faces - set - value decent
24:50 Recognition - tune - training data - minimum size of faces - effect - face recognized
26:35 Recognition - tune - results - minimum size of faces - set - value to big
27:35 Recognition - tune - results - minimum size of faces - effect - no face recognized
28:15 Recognition - tune - results - minimum size of faces - set - value decent
28:30 Recognition - tune - results - minimum size of faces - effect - face recognized
28:45 Recognition - tune - most similar faces - explanation
29:25 Insights - what files are written
29:55 Admin - activate all detectors and model
30:25 Insights - faces - activate all detectors and models
30:45 Insights - faces - activate all detectors and models - delete all result
31:00 Insights - faces - activate all detectors and models - start detection
31:35 Insights - faces - activate all detectors and models - effects - browser
31:50 Insights - faces - activate all detectors and models - effects - file statistics
32:40 Insights - history - model statistics - what combination of detector-model does the best job
33:35 Insights - history - model statistics - results - populate with results (by naming faces)
35:40 Insights - history - model statistics - results - order changes if names are set in browser
37:15 Insights - history - model statistics - results - see the results
37:45 Presets - experimental
38:15 Presets - reset - the defaults
38:59 The end

Find the best Settings for your Pictures

In theory the face recognition is quite accurate. What do real live examples show?
How to find the settings that work best for your pictures and how to apply them?



00:20 Mode Probe explained in short
02:40 Pages - help - links to the pages thresholds and probe
03:20 Page - thresholds - distance metrics
03:45 Page - probe - step-by-step instructions
04:10 Start the tutorial with a fresh system (no pictures yet)
04:25 Upload pictures
04:40 Mount cloud files via webDAV (to later save CPU and memory on the server)
06:15 Admin - block Python scripts on the server
06:40 Admin - activate all detectors, models, distance metrics
07:35 Settings - activate all detectors, models, distance metrics
09:10 Run the Python script on your computer to detect faces
09:50 Watch the log file to see what is going on
10:50 Addon - name faces
11:35 Upload more pictures - every picture contains just one face of the persons we named before
13:45 Upload more pictures - no picture does contain a face of the persons we named before
15:15 Start probe - via web browser (1st way)
15:40 Start prove - via terminal on your computer (2nd way)
15:45 Results of probe - file probe.csv - interpretations
21:50 Results of probe - file face_statistics.csv - look at internal data of the face recognition
24:30 Effectiveness of face recognition - real live
25:15 Effectiveness of face recognition - the theory
25:15 Why the feature "probe" was implemented - compare the theory against real live performance
26:20 Results of probe - find the combination that works best for your real live pictures
29:15 Apply the best combination - thresholds
30:00 Apply the best combination - settings
30:40 (Admin - activate the detectors, settings and distance metrics)
31:10 Remove results for detectors and models we do not want to use anymore (blocked)
31:50 (Admin - allow the python scripts)
32:20 Remove results for detectors and models we do not want to use anymore (again)
32:40 Addon - start the face recognition again using your optimized settings
33:20 File statistics.csv - show that unused detectors and models where removed
33:43 The end