EvGNN: An Event-driven Graph Neural Network Accelerator for Edge Vision.
Yufeng YangAdrian KneipCharlotte FrenkelPublished in: CoRR (2024)
Keyphrases
- event driven
- neural network
- weighted graph
- real time
- artificial neural networks
- information delivery
- computer vision
- disjoint paths
- edge weights
- graph representation
- undirected graph
- strongly connected
- edge detection
- publish subscribe
- genetic algorithm
- neural network model
- graph structure
- graph theory
- graph model
- bipartite graph
- vision system
- edge information
- structured data
- back propagation
- image processing
- parallel implementation
- directed graph
- markup language
- fault diagnosis
- pattern recognition
- neural network is trained