检索结果-内蒙古大学图书馆

Dynamic Graph Dismantling with Reinforcement Supervised Learning

学校读者我要写书评

暂无评论

Dynamic Graph Dismantling with Reinforcement Supervised Lear...

第42届中国控制会议

作者： Yumei Wang Chuancong Tang Hai-Tao Zhang School of Artificial Intelligence and Automation the Engineering Research Center of Autonomous Intelligent Unmanned Systemsthe Key Laboratory of Image Processing and Intelligent Controland the State Key Laboratory of Digital Manufacturing Equipment and TechnologyHuazhong University of Science and Technology

There are always some "key" nodes in a big complex network,which can joint the most connected *** to identify these nodes,finding a minimum set of nodes to attack for reducing the size of residual network's Largest Connected Component(LCC) to break up the original network,has become a research ***,a method for determining the"key" nodes based on reinforcement learning framework and supervised learning model is *** algorithm can not only utilize the dynamic exploration ability of reinforcement learning to collect a rich training dataset,but also take advantage of the characteristics that supervised learning is adaptive and has strong generalization ability to possess high efficiency and strong *** order to further improve the algorithm's performance,-greedy mechanism is used to explore more network *** experiment results show that given the same fraction of removed nodes,our algorithm can make the residual LCC smaller in various networks which is superior to the state-of-the-art algorithms in terms of effectiveness and generalization.

关键词：

Spatial-Temporal Constrained Pseudo-labeling for Unsupervised Person Re-identification via GCN Inference 18th

学校读者我要写书评

暂无评论

Spatial-Temporal Constrained Pseudo-labeling for Unsupervis...

18th International Forum of Digital Multimedia Communication, IFTC 2021

作者： Ling, Sen Yang, Hua Liu, Chuang Chen, Lin Zhao, Hongtian The Institute of Image Communication and Network Engineering Department of Electronic Engineering Shanghai Jiao Tong University Shanghai China Shanghai Key Laboratory of Digital Media Processing and Transmission Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9789811922657

Most existing unsupervised person re-identification (Re-ID) methods primarily depend on the cluster distance, and merely exploit the available source labeled data to assign pseudo labels for the unannotated data. Whereas, the cluster distance usually fails to adapt to different datasets due to the domain gap. Besides, learning exclusively from the source data can not generate accurate pseudo labels for the lack of the target data information. To address this problem, we propose to exploit the spatial-temporal constraints to facilitate the pseudo label generation process. Specifically, graphs for the labeled source data are constructed and the graph convolution network (GCN) is used to learn graph embeddings. Based on these graph embeddings, the likelihood of linkages between graph nodes is estimated and utilized to assign pseudo labels for the unlabeled data. Then, with the pseudo labels, a smoothed spatial-temporal probability distribution model is generated to amend the likelihood of linkages between graph nodes as well as correct the visual similarity scores for person Re-ID. Finally, we optimize the pseudo label assignment, feature extraction networks, and spatial-temporal model alternatively and iteratively to improve the person Re-ID performance. Comprehensive experiments demonstrate that the proposed method outperforms state-of-the-art methods. © 2022, Springer Nature Singapore Pte Ltd.

关键词： Graph theory

Disentangling Spatial and Temporal Learning for Efficient image-to-Video Transfer Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Qing, Zhiwu Zhang, Shiwei Huang, Ziyuan Zhang, Yingya Gao, Changxin Zhao, Deli Sang, Nong Key Laboratory of Image Processing Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology China Alibaba Group China ARC National University of Singapore Singapore

Recently, large-scale pre-trained language-image models like CLIP have shown extraordinary capabilities for understanding spatial contents, but naively transferring such models to video recognition still suffers from unsatisfactory temporal modeling capabilities. Existing methods insert tunable structures into or in parallel with the pre-trained model, which either requires back-propagation through the whole pre-trained model and is thus resource-demanding, or is limited by the temporal reasoning capability of the pre-trained structure. In this work, we present DiST, which disentangles the learning of spatial and temporal aspects of videos. Specifically, DiST uses a dual-encoder structure, where a pre-trained foundation model acts as the spatial encoder, and a lightweight network is introduced as the temporal encoder. An integration branch is inserted between the encoders to fuse spatio-temporal information. The disentangled spatial and temporal learning in DiST is highly efficient because it avoids the back-propagation of massive pre-trained parameters. Meanwhile, we empirically show that disentangled learning with an extra network for integration benefits both spatial and temporal understanding. Extensive experiments on five benchmarks show that DiST delivers better performance than existing state-of-the-art methods by convincing gaps. When pretraining on the large-scale Kinetics-710, we achieve 89.7% on Kinetics-400 with a frozen ViT-L model, which verifies the scalability of DiST. Codes and models can be found in https://***/alibaba-mmai-research/DiST. Copyright © 2023, The Authors. All rights reserved.

关键词： Backpropagation

VideoLCM: Video Latent Consistency Model

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Xiang Zhang, Shiwei Zhang, Han Liu, Yu Zhang, Yingya Gao, Changxin Sang, Nong Key Laboratory of Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology China Alibaba Group China Shanghai Jiao Tong University China

Consistency models have demonstrated powerful capability in efficient image generation and allowed synthesis within a few sampling steps, alleviating the high computational cost in diffusion models. However, the consi...

关键词：

A Physiological Signal Emotion Recognition Method Based on Domain Adaptation and Incremental Learning

学校读者我要写书评

暂无评论

A Physiological Signal Emotion Recognition Method Based on D...

Cyber-Energy Systems and Intelligent Energy (ICCSIE), International Conference on

作者： Junnan Li Xiaoping Wang the School of Artificial Intelligence and Automation and the Key Laboratory of Image Processing and Intelligent Control of Education Ministry of China Huazhong University of Science and Technology Wuhan P. R. China

Temporal concept shift (TCS) is an unavoidable problem in physiological signal-based emotion recognition tasks, i.e., the data distribution of physiological signals is constantly changing over time, which gradually degrades the model accuracy. To this end, we propose a method based on a combination of domain adaptation and incremental learning to reduce the impact of temporal concept drift. In this paper, domain adaptation is used to reduce the distribution differences and incremental learning is used to prevent the learned knowledge from being forgotten. Finally, we validate the effectiveness of our approach on two real datasets.

关键词：

3D Cinemagraphy from a Single image

学校读者我要写书评

暂无评论

3D Cinemagraphy from a Single Image

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Xingyi Li Zhiguo Cao Huiqiang Sun Jianming Zhang Ke Xian Guosheng Lin Key Laboratory of Image Processing and Intelligent Control Ministry of Education School of Artificial Intelligence and Automation Huazhong University of Science and Technology S-Lab Nanyang Technological University Adobe Research

We present 3D Cinemagraphy, a new technique that mar-ries 2D image animation with 3D photography. Given a single still image as input, our goal is to generate a video that contains both visual content animation and camera motion. We empirically find that naively combining existing 2D image animation and 3D photography methods leads to obvious artifacts or inconsistent animation. Our key insight is that representing and animating the scene in 3D space offers a natural solution to this task. To this end, we first convert the input image into feature-based layered depth images using predicted depth values, followed by unprojecting them to a feature point cloud. To animate the scene, we perform motion estimation and lift the 2D motion into the 3D scene flow. Finally, to resolve the problem of hole emer-gence as points move forward, we propose to bidirectionally displace the point cloud as per the scene flow and synthe-size novel views by separately projecting them into target image planes and blending the results. Extensive experiments demonstrate the effectiveness of our method. A user study is also conducted to validate the compelling rendering results of our method.

关键词：

Point-and-Shoot All-in-Focus Photo Synthesis from Smartphone Camera Pair

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Luo, Xianrui Peng, Juewen Zhao, Weiyue Xian, Ke Lu, Hao Cao, Zhiguo The Key Laboratory of Image Processing and Intelligent Control Ministry of Education The School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China The S-Lab for Advanced Intelligence Nanyang Technological University Singapore

All-in-Focus (AIF) photography is expected to be a commercial selling point for modern smartphones. Standard AIF synthesis requires manual, time-consuming operations such as focal stack compositing, which is unfriendly to ordinary people. To achieve point-and-shoot AIF photography with a smartphone, we expect that an AIF photo can be generated from one shot of the scene, instead of from multiple photos captured by the same camera. Benefiting from the multi-camera module in modern smartphones, we introduce a new task of AIF synthesis from main (wide) and ultra-wide cameras. The goal is to recover sharp details from defocused regions in the main-camera photo with the help of the ultra-wide-camera one. The camera setting poses new challenges such as parallax-induced occlusions and inconsistent color between cameras. To overcome the challenges, we introduce a predict-and-refine network to mitigate occlusions and propose dynamic frequency-domain alignment for color correction. To enable effective training and evaluation, we also build an AIF dataset with 2686 unique scenes. Each scene includes two photos captured by the main camera, one photo captured by the ultra-wide camera, and a synthesized AIF photo. Results show that our solution, termed EasyAIF, can produce high-quality AIF photos and outperforms strong baselines quantitatively and qualitatively. For the first time, we demonstrate point-and-shoot AIF photo synthesis successfully from main and ultra-wide cameras. Copyright © 2023, The Authors. All rights reserved.

关键词： Cameras

Alignment-Based Adversarial Training (ABAT) for Improving the Robustness and Accuracy of EEG-Based BCIs

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Chen, Xiaoqing Wang, Ziwei Wu, Dongrui The Key Laboratory of the Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen China

Machine learning has achieved great success in electroencephalogram (EEG) based brain-computer interfaces (BCIs). Most existing BCI studies focused on improving the decoding accuracy, with only a few considering the adversarial security. Although many adversarial defense approaches have been proposed in other application domains such as computer vision, previous research showed that their direct extensions to BCIs degrade the classification accuracy on benign samples. This phenomenon greatly affects the applicability of adversarial defense approaches to EEG-based BCIs. To mitigate this problem, we propose alignment-based adversarial training (ABAT), which performs EEG data alignment before adversarial training. Data alignment aligns EEG trials from different domains to reduce their distribution discrepancies, and adversarial training further robustifies the classification boundary. The integration of data alignment and adversarial training can make the trained EEG classifiers simultaneously more accurate and more robust. Experiments on five EEG datasets from two different BCI paradigms (motor imagery classification, and event related potential recognition), three convolutional neural network classifiers (EEGNet, ShallowCNN and DeepCNN) and three different experimental settings (offline within-subject cross-block/-session classification, online cross-session classification, and pre-trained classifiers) demonstrated its effectiveness. It is very intriguing that adversarial attacks, which are usually used to damage BCI systems, can be used in ABAT to simultaneously improve the model accuracy and robustness. © 2024, CC BY.

关键词： Adversarial machine learning

Statistical Feasibility Robust Optimization for Hydrogen-Data Center Microgrid Operation

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Duan, Juntao Li, Yuanzheng Li, Yang Liu, Zhiwei Zhao, Yong School of Artificial Intelligence and Automation Key Laboratory on Image Information Processing and Intelligent Control of Ministry of Education Huazhong University of Science and Technology Wuhan430074 China Key Laboratory of Modern Power System Simulation and Control & Renewable Energy Technology Ministry of Education Northeast Electric Power University Jilin132012 China

The increasing processing demands of information technology (IT) workloads cause the booming of data centers (DCs) in recent years. The servers in the DC emit much heat when processing the IT workloads and thus the DC temperature rises. The power consumption when controlling the DC temperature using the air conditioning (A/C) of the DC is high, which leads to expensive power bills. To reduce the power cost, using renewable energy (RE) to power DC has raised much attention in recent years. However, the intermittency of RE and the uncertainties of outside temperature pose operational challenges. Therefore, in this paper, a data-driven statistical-feasibility-based robust rolling optimization framework for the operation of hydrogen-DC microgrid (H2-DCMG) is proposed to control the DC temperature considering these uncertainties and the intermittency of RE. In the framework, the hydrogen energy storage system (HESS) is introduced to deal with the intermittency of RE, and robust optimization (RO) is adopted to deal with the uncertainty of outside temperature. The statistical feasibility is introduced into the framework to decrease the conservatism of RO, through relaxing the DC temperature constraints without harming normal DC operations. Moreover, we propose the Statistical-Guarantee-based Vertex Link (SGVL) algorithm to construct a polyhedron uncertain set of outside temperatures that can ensure the results of dispatch are economical and satisfy statistical feasibility. The constructed polyhedron uncertain set is linear, which reduces computational complexity in solving. A case study based on the real-world temperature dataset verifies the superiority of the SGVL algorithm on the aspect of the economic cost. It further validates the effectiveness of the proposed robust optimization framework for DC temperature control. © 2024, The Authors. All rights reserved.

关键词： Air conditioning