Neural network learns to identify group sizes without knowledge of numbers

After feeding the network 51,800 images, where each was a unique layout of rectangles of various sizes, the researchers found that the new images generated by the system began to demonstrate an awareness of the relative size of different groups without having to perform any counting. Next, to demonstrate, ANS, the team fed the system another program that allowed the system to compare different groups that it had seen during the first run and found, based on new images generated, that the system was able to make educated guesses about which was bigger or smaller; Teaching computer systems to learn to use ANS is but one step towards creating machines that think rather than simply crunch numbers for us, and the hope is that one day, such systems can be put into robots to make them as useful as those weve seen in movies for decades.

Author: Publication Date: Crawl Date: Source: Topic: Contributor: Type: Language: Format:
January 23, 2012 January 30, 2012 via Google News CognitiveScience MachineLearning Vision NewsFinder Text English HTML

Whitelist words:

  1. artifici intellig: 1 occurrence
  2. cognit scienc: 1 occurrence
  3. neural network: 1 occurrence

Categorization (need >= 0.5 to match):

  1. AIOverview: 0.00715586
  2. Agents: 0.00677101
  3. Applications: 0.267983
  4. CognitiveScience: 0.885378
  5. Education: 0.29747
  6. Ethics: 0.0033112
  7. Games: 0.00444924
  8. History: 0.00952492
  9. Interfaces: 0.00854047
  10. MachineLearning: 0.921872
  11. NaturalLanguage: 0.00098136
  12. Philosophy: 0.0096514
  13. Reasoning: 0.406807
  14. Representation: 0.0092676
  15. Robots: 0.0417772
  16. ScienceFiction: 0.0192979
  17. Speech: 0.0265272
  18. Systems: 0.0309579
  19. Vision: 0.605431

Duplicates (threshold=0.17):

  1. (sim=0.373) 5746: Neural network gets an idea of number without counting
  2. (sim=0.373) 5779: Neural network gets an idea of number without counting


  1. Published
AAAI   Recent Changes   Edit   History   Print   Contact Us
Page last modified on January 30, 2012, at 06:09 AM