My collection of deep learning solutions based on dynamic network configuration and attention mechanism. Awesome list of inspiration is here
Article | Year | Affiliation | Code | Smol Description | Key Words |
---|---|---|---|---|---|
Dynamic Algorithm Configuration:Foundation of a New Meta-Algorithmic Framework [blog] | 2020 | Germany | code | Hyperparameter optimization | |
BOHB: Robust and Efficient Hyperparameter Optimization at Scale [blog] | 2018 | Germany | code | Hyperparameter optimization | |
Not All Attention Is Needed: Gated Attention Network for Sequence Data [blog] | 2019 | Hong-Kong, Amazon | -/- | Attention mechanism, dynamic network configuration, sequential models, NLP, text classification | |
Multi-Dimension Modulation for Image Restoration with Dynamic Controllable Residual Learning [review] | 2019 | China | code | Image restoration, interactive multi-dimension modulation | |
MixPath: A Unified Approach for One-shot Neural Architecture Search | 2020 | China, Xiaomi | code | supernet, multi-path search space | |
Graph Attention Networks [blog] | 2018 | Europe, Canada | code | graph, attention mechanism, graph-structured data | |
Fully Convolutional Network with Multi-Step Reinforcement Learningfor Image Processing [review} | 2018 | Japan | -/- | RL, pixel-wise agents, image restoration | |
Balanced One-shot Neural Architecture Optimization | 2019 | China, Microsoft Asia | -/- | supernet, neural architecture search |
Interesting popular articles: