• Publications
  • Influence
FasterSeg: Searching for Faster Real-time Semantic Segmentation
TLDR
We present FasterSeg, an automatically designed semantic segmentation network with not only state-of-the-art performance but also faster speed than current methods. Expand
  • 27
  • 4
  • PDF
AugFPN: Improving Multi-Scale Feature Learning for Object Detection
TLDR
We introduce a new feature pyramid architecture named AugFPN to address the design defects of feature pyramid in Faster R-CNN. Expand
  • 15
  • 3
  • PDF
RENAS: Reinforced Evolutionary Neural Architecture Search
TLDR
We propose the Reinforced Evolutionary Neural Architecture Search (RENAS), which is an evolutionary method with reinforced mutation for neural architecture exploration. Expand
  • 29
  • 2
  • PDF
Attention-based recurrent neural networks for accurate short-term and long-term dissolved oxygen prediction
TLDR
In this paper, we study the effectiveness of attention-based recurrent neural networks (RNN) on short-term prediction (including about 1h, 2 h) and long-term Prediction (including 8h, 24h and 48h) of dissolved oxygen. Expand
  • 11
  • 2
An enhanced Bacterial Foraging Optimization and its application for training kernel extreme learning machine
TLDR
An enhanced BFO with chaotic chemotaxis step length, Gaussian mutation and chaotic local search (CCGBFO) is proposed for overcoming the existing weakness of original BFO. Expand
  • 100
  • 1
Saliency Detection via the Improved Hierarchical Principal Component Analysis Method
TLDR
A visual saliency detection algorithm based on Hierarchical Principal Component Analysis (HPCA) has been proposed in the paper. Expand
  • 26
  • 1
  • PDF
A nature-inspired multi-objective optimisation strategy based on a new reduced space searching algorithm for the design of alloy steels
TLDR
In this paper, a salient search and optimisation algorithm based on a new reduced space searching strategy is presented. Expand
  • 31
  • 1
  • PDF
Self-Ensembling Attention Networks: Addressing Domain Shift for Semantic Segmentation
TLDR
We propose self-ensembling attention networks to reduce the domain gap between different datasets to address domain shift for semantic segmentation. Expand
  • 14
  • 1
Video Object Detection with Locally-Weighted Deformable Neighbors
TLDR
We propose LWDN (Locally-Weighted Deformable Neighbors) for video object detection without utilizing time-consuming optical flow extraction networks. Expand
  • 10
  • 1
  • PDF