FFCL: Forward-Forward Net with Cortical Loops, Training and Inference on Edge Without Backpropogation.
Ali KarkehabadiHouman HomayounAvesta SasanPublished in: ACM Great Lakes Symposium on VLSI (2024)
Keyphrases
- forward and backward
- bayesian networks
- databases
- bi directional
- supervised learning
- edge detection
- inference process
- information systems
- image segmentation
- database
- receptive fields
- online learning
- cortical surface
- training phase
- biologically inspired
- edge detector
- training examples
- probability distribution
- machine learning
- data sets