G
Gao Huang
Researcher at Tsinghua University
Publications - 164
Citations - 43663
Gao Huang is an academic researcher from Tsinghua University. The author has contributed to research in topics: Computer science & Feature (computer vision). The author has an hindex of 37, co-authored 124 publications receiving 26697 citations. Previous affiliations of Gao Huang include Cornell University & University of Science and Technology of China.
Papers
More filters
Posted Content
An empirical study on evaluation metrics of generative adversarial networks
TL;DR: This paper comprehensively investigates existing sample-based evaluation metrics for GANs and observes that kernel Maximum Mean Discrepancy and the 1-Nearest-Neighbor (1-NN) two-sample test seem to satisfy most of the desirable properties, provided that the distances between samples are computed in a suitable feature space.
Proceedings ArticleDOI
Gradient boosted feature selection
TL;DR: Gradient Boosted Feature Selection (GBFS) as mentioned in this paper is a feature selection algorithm that is based on a modification of gradient boosted trees and is shown to match or outperform other state of the art feature selection algorithms.
Proceedings Article
Supervised word mover's distance
TL;DR: This paper proposes an efficient technique to learn a supervised metric, which it is called the Supervised-WMD (S-W MD) metric, and provides an arbitrarily close approximation of the original WMD distance that results in a practical and efficient update rule.
Posted Content
Memory-Efficient Implementation of DenseNets
TL;DR: This technical report introduces strategies to reduce the memory consumption of DenseNets during training by strategically using shared memory allocations, and reduces the memory cost for storing feature maps from quadratic to linear.
Posted Content
Multi-Scale Dense Convolutional Networks for Efficient Prediction.
TL;DR: A new convolutional neural network architecture with the ability to adapt dynamically to computational resource limits at test time and substantially improves the state-of-the-art in both settings is introduced.