Added: 3 years ago
From: RoboThought
Views: 878
Sort by time | Sort by thread (beta)

Link to this comment:

Share to:

All Comments (2)

Sign In or Sign Up now to post a comment!
  • I had the chance to witness it performing live recognition. It's abolutely amazing for everything it can do. And it's just the beginning...

  • Hi, Here are the summarized specifications: - 1024 neurons, expandable by cascading chips - Choose Radial Basis Function or KNN non-linear classifiers - Automatic model generator in learning mode - Pattern data max length: 256 - Pattern data resolution: 1 byte - Category values: 32768 (15-bit) - Distance resolution: 65536 (16-bit) - Distance norm: L1/LSup - Adjustable min/max influence fields - Max # sub-networks or contexts: 128 (7-bit)
  • Fascinating. May I ask how many neurons are on the chip and how many objects it can be trained to recognize simultaneously? Have you thought of implementing fiber optic channels between the neurons instead of etched wired connections?

Loading...
Alert icon
0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more