Deep Learning on Shallow Hardware
Adnan Siddiqui
@adnans
None (we just use email)
https://github.com/deepseattle/shallow.hardware.projects
In this project, we will be implementing a facial recognition model on live video stream on shallow (low spec/embedded) hardware.
The Labeled Faces in the Wild (LFW) data set at http://vis-www.cs.umass.edu/lfw/ has been proposed by the team members for this project.
The hardware choices proposed so far are:
- Raspberry Pi
- ASUS Tinker Board
This is hardware for implementing the production model. Training will still be done on high power GPU/CPU machines.
TBD (likely every two to three weeks in-person as well as web conferenc)
Contact the project leader (Adnan) via emai at adnans@seealgo.com to join the project team