Corpus ID: 49342623

Resource-Efficient Neural Architect

  title={Resource-Efficient Neural Architect},
  author={Yanqi Zhou and S. Ebrahimi and Sercan {\"O}. Arik and Haonan Yu and Hairong Liu and Greg Diamos},
  • Yanqi Zhou, S. Ebrahimi, +3 authors Greg Diamos
  • Published 2018
  • Computer Science
  • ArXiv
  • Neural Architecture Search (NAS) is a laborious process. Prior work on automated NAS targets mainly on improving accuracy, but lacks consideration of computational resource use. We propose the Resource-Efficient Neural Architect (RENA), an efficient resource-constrained NAS using reinforcement learning with network embedding. RENA uses a policy network to process the network embeddings to generate new configurations. We demonstrate RENA on image recognition and keyword spotting (KWS) problems… CONTINUE READING
    33 Citations

    Figures, Tables, and Topics from this paper

    Layer-compensated Pruning for Resource-constrained Convolutional Neural Networks
    • 31
    • PDF
    RC-DARTS: Resource Constrained Differentiable Architecture Search
    • 5
    • PDF
    InstaNAS: Instance-aware Neural Architecture Search
    • 14
    • PDF
    Multi-Objective Neural Architecture Search for Light-Weight Model
    • 2
    Inductive Transfer for Neural Architecture Optimization
    • 7
    • PDF
    Single-Path Mobile AutoML: Efficient ConvNet Design and NAS Hyperparameter Optimization
    • 9
    • PDF
    Learnable Embedding Space for Efficient Neural Architecture Compression
    • 21
    • PDF
    Graph HyperNetworks for Neural Architecture Search
    • 94
    • PDF
    Weight-Sharing Neural Architecture Search: A Battle to Shrink the Optimization Gap
    • 5
    • PDF


    Efficient Neural Architecture Search via Parameter Sharing
    • 1,079
    • PDF
    Multi-objective Architecture Search for CNNs
    • 48
    Hierarchical Representations for Efficient Architecture Search
    • 474
    • PDF
    Designing Neural Network Architectures using Reinforcement Learning
    • 730
    • PDF
    N2N Learning: Network to Network Compression via Policy Gradient Reinforcement Learning
    • 92
    • PDF
    DeepArchitect: Automatically Designing and Training Deep Architectures
    • 127
    • PDF
    Learning both Weights and Connections for Efficient Neural Network
    • 2,882
    • PDF
    Large-Scale Evolution of Image Classifiers
    • 790
    • PDF
    Regularized Evolution for Image Classifier Architecture Search
    • 1,049
    • PDF
    Rethinking the Inception Architecture for Computer Vision
    • 10,266
    • PDF