Feed-Forward On-Edge Fine-Tuning Using Static Synthetic Gradient Modules.
Robby NevenMarian VerhelstTinne TuytelaarsToon GoedeméPublished in: ECCV Workshops (5) (2020)
Keyphrases
- feed forward
- fine tuning
- edge detection
- back propagation
- image gradient
- artificial neural networks
- neural nets
- gradient direction
- gradient information
- step edges
- recurrent neural networks
- viable alternative
- orientation histogram
- edge orientation
- neural network
- fine tune
- biologically plausible
- edge detector
- edge map
- gradient field
- edge information
- gradient magnitude
- hidden layer
- fine tuned
- error back propagation
- neural architecture
- feed forward neural networks
- zero crossing
- visual cortex
- spiking neural networks
- activation function
- artificial neural
- multiscale
- mumford shah model
- weighted graph
- optical flow
- edge direction
- knowledge base
- neuron model
- artificial intelligence