StyleNAS: An Empirical Study of Neural Architecture Search to Uncover Surprisingly Fast End-to-End Universal Style Transfer Networks.
Jie AnHaoyi XiongJinwen MaJiebo LuoJun HuanPublished in: CoRR (2019)
Keyphrases
- end to end
- neural architecture
- high bandwidth
- wireless ad hoc networks
- internet protocol
- search space
- admission control
- ad hoc networks
- neural network
- scalable video
- differentiated services
- congestion control
- feed forward
- text localization and recognition
- real time
- application layer
- activation function
- computer networks
- artificial neural networks
- genetic algorithm
- real world