no code implementations • 19 Apr 2022 • Wei Chen, Jie Zhao, Wan-Lei Zhao, Song-Yuan Wu
In this paper, a single-stage monocular 3D object detection model is proposed.
no code implementations • 2 Apr 2022 • Hui Wang, Yong Wang, Wan-Lei Zhao
For the graph construction, a two-stage graph diversification scheme is proposed, which makes a good trade-off between the efficiency and reachability for the search procedure that builds upon it.
no code implementations • 10 Mar 2022 • Gao-Dong Liu, Wan-Lei Zhao, Jie Zhao
Existing solutions either retrain the model from scratch or require the replay of old samples during the training.
1 code implementation • CIKM '21: Proceedings of the 30th ACM International Conference on Information & Knowledge Management 2021 • Hui Wang, Wan-Lei Zhao, Xiangxiang Zeng, Jianye Yang
In this paper, NN-Descent has been redesigned to adapt to the GPU architecture.
no code implementations • 27 Aug 2021 • Shi-Ying Lan, Run-Qing Chen, Wan-Lei Zhao
Anomaly detection on time series is a fundamental task in monitoring the Key Performance Indicators (KPIs) of IT systems.
no code implementations • 11 Jul 2021 • Yi-Geng Hong, Hui-Chu Xiao, Wan-Lei Zhao
In this paper, a self-paced learning framework is proposed to achieve accurate object localization on the rank list returned by instance search.
1 code implementation • 29 Mar 2021 • Hui Wang, Wan-Lei Zhao, Xiangxiang Zeng
NN-Descent was proposed as an effective solution for the graph construction problem.
1 code implementation • 19 May 2020 • Wan-Lei Zhao, Run-Qing Chen, Hui Ye, Chong-Wah Ngo
This optimization procedure converges faster to a better local minimum over k-means and many of its variants.
1 code implementation • 24 Apr 2020 • Chang-Hui Liang, Wan-Lei Zhao, Run-Qing Chen
In this paper, a dynamic sampling strategy is proposed to organize the training pairs in an easy-to-hard order to feed into the network.
no code implementations • 1 Feb 2020 • Hui-Chu Xiao, Wan-Lei Zhao, Jie Lin, Chong-Wah Ngo
Due to the lack of proper mechanism in locating instances and deriving feature representation, instance search is generally only effective for retrieving instances of known object categories.
no code implementations • 9 Oct 2019 • Run-Qing Chen, Guang-Hui Shi, Wan-Lei Zhao, Chang-Hui Liang
Status prediction and anomaly detection are two fundamental tasks in automatic IT systems monitoring.
1 code implementation • 2 Aug 2019 • Wan-Lei Zhao, Hui Wang, Peng-Cheng Lin, Chong-Wah Ngo
Unfortunately, a closely related issue of how to merge two existing k-NN graphs has been overlooked.
no code implementations • 3 Apr 2019 • Peng-Cheng Lin, Wan-Lei Zhao
Recently, graph based nearest neighbor search gets more and more popular on large-scale retrieval tasks.
1 code implementation • 10 Jun 2018 • Yu Zhan, Wan-Lei Zhao
In addition, the proposed enhancement on the network structure also shows superior performance on the instance segmentation task.
1 code implementation • 12 Apr 2018 • Cheng-Hao Deng, Wan-Lei Zhao
In order to detect clusters in arbitrary shapes, a novel and generic solution based on boundary erosion is proposed.
no code implementations • 9 Apr 2018 • Wan-Lei Zhao, Hui Wang, Chong-Wah Ngo
On the one hand, the approximate k-nearest neighbor graph construction is treated as a search task.
no code implementations • 4 May 2017 • Cheng-Hao Deng, Wan-Lei Zhao
In the k-means iteration, each data sample is only compared to clusters that its nearest neighbors reside.
no code implementations • 28 Apr 2017 • Guanjun Guo, Hanzi Wang, Wan-Lei Zhao, Yan Yan, Xuelong. Li
Based on the new Cohesion Measurement, a novel object discovery method is proposed to discover objects latent in an image by utilizing the eigenvectors of the affinity matrix.
no code implementations • 30 Jan 2017 • Wan-Lei Zhao, Jie Yang, Cheng-Hao Deng
In this paper, a scalable solution based on hill-climbing strategy with the support of k-nearest neighbor graph (kNN) is presented.
no code implementations • 8 Oct 2016 • Wan-Lei Zhao, Cheng-Hao Deng, Chong-Wah Ngo
The performance of k-means has been enhanced from different perspectives over the years.