Zobrazeno 1 - 10
of 142
pro vyhledávání: '"HE Haoyang"'
Publikováno v:
Chengshi guidao jiaotong yanjiu, Vol 27, Iss 9, Pp 236-241 (2024)
Objective Data interconnection and interaction are realized through ATP (automatic train protection) system in most of the existing train speed and distance measuring technologies. However, the implementation process is complicated, and the train pos
Externí odkaz:
https://doaj.org/article/9dfe353d5aa4457a989b8f5abd46cb53
Autor:
Cai, Yuxuan, Zhang, Jiangning, He, Haoyang, He, Xinwei, Tong, Ao, Gan, Zhenye, Wang, Chengjie, Bai, Xiang
The success of Large Language Models (LLM) has led researchers to explore Multimodal Large Language Models (MLLM) for unified visual and linguistic understanding. However, the increasing model size and computational complexity of MLLM limit their use
Externí odkaz:
http://arxiv.org/abs/2410.16236
This study explores the recently proposed challenging multi-view Anomaly Detection (AD) task. Single-view tasks would encounter blind spots from other perspectives, resulting in inaccuracies in sample-level prediction. Therefore, we introduce the \te
Externí odkaz:
http://arxiv.org/abs/2407.11935
Autor:
Zhang, Jiangning, He, Haoyang, Gan, Zhenye, He, Qingdong, Cai, Yuxuan, Xue, Zhucun, Wang, Yabiao, Wang, Chengjie, Xie, Lei, Liu, Yong
Visual anomaly detection aims to identify anomalous regions in images through unsupervised learning paradigms, with increasing application demand and value in fields such as industrial inspection and medical lesion detection. Despite significant prog
Externí odkaz:
http://arxiv.org/abs/2406.03262
Autor:
He, Qingdong, Zhang, Jiangning, Peng, Jinlong, He, Haoyang, Li, Xiangtai, Wang, Yabiao, Wang, Chengjie
Transformers have revolutionized the point cloud learning task, but the quadratic complexity hinders its extension to long sequence and makes a burden on limited computational resources. The recent advent of RWKV, a fresh breed of deep sequence model
Externí odkaz:
http://arxiv.org/abs/2405.15214
Autor:
He, Haoyang, Bai, Yuhu, Zhang, Jiangning, He, Qingdong, Chen, Hongxu, Gan, Zhenye, Wang, Chengjie, Li, Xiangtai, Tian, Guanzhong, Xie, Lei
Recent advancements in anomaly detection have seen the efficacy of CNN- and transformer-based approaches. However, CNNs struggle with long-range dependencies, while transformers are burdened by quadratic computational complexity. Mamba-based models,
Externí odkaz:
http://arxiv.org/abs/2404.06564
Autor:
He, Haoyang, Zhang, Jiangning, Chen, Hongxu, Chen, Xuhai, Li, Zhishan, Chen, Xu, Wang, Yabiao, Wang, Chengjie, Xie, Lei
Reconstruction-based approaches have achieved remarkable outcomes in anomaly detection. The exceptional image reconstruction capabilities of recently popular diffusion models have sparked research efforts to utilize them for enhanced reconstruction o
Externí odkaz:
http://arxiv.org/abs/2312.06607
Autor:
Zhang, Jiangning, He, Haoyang, Chen, Xuhai, Xue, Zhucun, Wang, Yabiao, Wang, Chengjie, Xie, Lei, Liu, Yong
Large Multimodal Model (LMM) GPT-4V(ision) endows GPT-4 with visual grounding capabilities, making it possible to handle certain tasks through the Visual Question Answering (VQA) paradigm. This paper explores the potential of VQA-oriented GPT-4V in t
Externí odkaz:
http://arxiv.org/abs/2311.02612
Autor:
Chen, Xuhai, Zhang, Jiangning, Tian, Guanzhong, He, Haoyang, Zhang, Wuhao, Wang, Yabiao, Wang, Chengjie, Liu, Yong
This paper considers zero-shot Anomaly Detection (AD), performing AD without reference images of the test objects. We propose a framework called CLIP-AD to leverage the zero-shot capabilities of the large vision-language model CLIP. Firstly, we reint
Externí odkaz:
http://arxiv.org/abs/2311.00453
Autor:
He, Haoyang
Model-based approaches are becoming increasingly popular in the field of offline reinforcement learning, with high potential in real-world applications due to the model's capability of thoroughly utilizing the large historical datasets available with
Externí odkaz:
http://arxiv.org/abs/2305.03360