GhostNet: More Features From Cheap Operations
- Kai Han, Yunhe Wang, Qi Tian, Jianyuan Guo, Chunjing Xu, Chang Xu
- Computer ScienceComputer Vision and Pattern Recognition
- 27 November 2019
A novel Ghost module is proposed to generate more feature maps from cheap operations based on a set of intrinsic feature maps to generate many ghost feature maps that could fully reveal information underlying intrinsic features.
Transformer in Transformer
- Kai Han, An Xiao, E. Wu, Jianyuan Guo, Chunjing Xu, Yunhe Wang
- Computer ScienceNeural Information Processing Systems
- 27 February 2021
It is pointed out that the attention inside these local patches are also essential for building visual transformers with high performance and a new architecture, namely, Transformer iN Transformer (TNT), is explored.
Greedy Hash: Towards Fast Optimization for Accurate Hash Coding in CNN
- Shupeng Su, Chao Zhang, Kai Han, Yonghong Tian
- Computer ScienceNeural Information Processing Systems
- 2018
This work adopts the greedy principle to tackle this NP hard problem by iteratively updating the network toward the probable optimal discrete solution in each iteration, and provides a new perspective to visualize and understand the effectiveness and efficiency of the algorithm.
Open-Set Recognition: A Good Closed-Set Classifier is All You Need
- S. Vaze, Kai Han, A. Vedaldi, Andrew Zisserman
- Computer ScienceInternational Conference on Learning…
- 12 October 2021
The ‘Semantic Shift Benchmark’ (SSB) is presented, which better respects the task of detecting semantic novelty, in contrast to other forms of distribution shift also considered in related sub-fields, such as out-of-distribution detection.
Co-Evolutionary Compression for Unpaired Image Translation
- Han Shu, Yunhe Wang, Chang Xu
- Computer ScienceIEEE International Conference on Computer Vision
- 25 July 2019
A novel co-evolutionary approach for reducing their memory usage and FLOPs simultaneously and synergistically optimized for investigating the most important convolution filters iteratively is developed.
CMT: Convolutional Neural Networks Meet Vision Transformers
- Jianyuan Guo, Kai Han, Yunhe Wang
- Computer ScienceComputer Vision and Pattern Recognition
- 13 July 2021
A new transformer based hybrid network is proposed by taking advantage of transformers to capture long-range dependencies, and of CNNs to extract local information, obtaining much better trade-off for accuracy and efficiency than previous CNN-based and transformer-based models.
Distilling Object Detectors via Decoupled Features
- Jianyuan Guo, Kai Han, Chang Xu
- Computer ScienceComputer Vision and Pattern Recognition
- 26 March 2021
This paper presents a novel distillation algorithm via decoupled features (DeFeat) for learning a better student detector that is able to surpass the state-of-the-art distillation methods for object detection.
Autoencoder Inspired Unsupervised Feature Selection
- Kai Han, Yunhe Wang, Chao Zhang, C. Li, Chao Xu
- Computer ScienceIEEE International Conference on Acoustics…
- 23 October 2017
Compared to traditional feature selection methods, AEFS can select the most important features by excavating both linear and nonlinear information among features, which is more flexible than the conventional self-representation method for unsupervised feature selection with only linear assumptions.
Attribute-Aware Attention Model for Fine-grained Representation Learning
- Kai Han, Jianyuan Guo, Chao Zhang, Mingjian Zhu
- Computer ScienceACM Multimedia
- 15 October 2018
A novel Attribute-Aware Attention Model is proposed, which can learn local attribute representation and global category representation simultaneously in an end-to-end manner and contains more intrinsic information for image recognition instead of the noisy and irrelevant features.
ReNAS: Relativistic Evaluation of Neural Architecture Search
- Yixing Xu, Yunhe Wang, Chang Xu
- Computer ScienceComputer Vision and Pattern Recognition
- 30 September 2019
This paper proposes a relativistic architecture performance predictor in NAS (ReNAS), encoding neural architectures into feature tensors, and further refining the representations with the predictor, to determine which architecture would perform better instead of accurately predict the absolute architecture performance.
...
...