Tags
Let's Face it.
This new article on IEEE Computers Magazine is a very good read, giving good background but also pointing to future directions. Very insightful. Highly recommended.
Here are a few directions I noted:
- on a psychological side, what is the role of top-down expectation? Will it help fast identification in a ‘many faces’ scenario (see my last post)? The familiarity of the face counts?
- how can a camera network do face recognition? The multi-view fusion should be done on the face image, or on face recognition results?
Someone volunteers a 411 talk on this subject?
