DTNN: Energy-efficient Inference with Dendrite Tree Inspired Neural Networks for Edge Vision Applications.
Tao LuoWai Teng TangMatthew Kay Fei LeeChuping QuWeng-Fai WongRick Siow Mong GohPublished in: CoRR (2021)
Keyphrases
- efficient inference
- neural network
- graph structure
- probabilistic inference
- conditional random fields
- fully connected
- markov random field
- junction tree
- tree structure
- computer vision
- human pose estimation
- artificial neural networks
- hidden variables
- approximate inference
- back propagation
- exact inference
- graphical models
- neural network model
- factor graphs
- linear models
- feed forward
- bayesian networks
- structured prediction
- neural nets
- markov networks
- probabilistic model
- index structure
- message passing