Photo Archive Applying Offline Facial Recognition Technology

I&T Solution Photo Archive Applying Offline Facial Recognition Technology
(REF: S-0944)
Trial Project
Solution Feature
  • Effective auditing and fast retrieval of a large person photos database.
  • Apply cutting edge Convolution Neural Network deep learning algorithms to train the facial recognition AI engine.
  • Tag and bookmark VIPs photos automatically or manually with selection criteria using interactive user interface.
  • Support searching VIP photos by date and event name which appear in the folder names and the photos themselves.
  • Auto check of duplicated photos inside the data management system to avoid memory usage redundancy.
Trial Application and Expected Outcome
  • Gather training data to start training the facial recognition AI engine.
  • Evaluate the performance by accuracy (compare recognition result with ground truth) and inference time (how fast it is returned) [Internal trial].
  • Setup and test a docker container with web portal and photo database query functions [Internal trial].
  • Deploy the web portal with facial recognition AI engine as a docker image to the stakeholder's on-premises server [Trial run].
  • Test the web server by uploading the database photos for users to search and make photo albums using person photos [Trial run].
Additional Solution Information 2_Proposal-PhotoArchiveFacialRecog_2021-03-01_v3.pdf
Info on I&T Solution Provider
Solution Provider:High Tech Technology Limited
Address:Unit 713, 7/F, Building 12W, No. 12 Science Park West Avenue, Hong Kong Science Park, Shatin, N.T., Hong Kong.
Contact Person:Steve Chim
Position:Staff Engineer
Tel:+852 3619 5374
Email: stevechim@hightt.com
Webpage: www.hightt.com

For details of the above I&T solution, please contact the I&T solution provider.