Haicheng Qu

Liaoning Technical University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Wanjun Liu

Liaoning Technical University

Jitao Qin

Liaoning Technical University

Junping Zhang

Harbin Institute of Technology

Zhouhan Lin

Shanghai Jiao Tong University

Yimeng Shan

Chinese Academy of Agricultural Sciences

Xuejian Liang

Harbin Institute of Technology

Ye Zhang

Beijing Health Vocational College

Hao Chen

Changzhou University

Lamei Liu

Liaoning Technical University

Haodong Li

Liaoning Technical University

Cooperative Institutions

Liaoning Technical University

Chinese Academy of Sciences

Harbin Institute of Technology

University of Chinese Academy of Sciences

Shanghai Jiao Tong University

Beijing Institute of Technology

National University of Singapore

Shenzhen University

Institute of Software

Hong Kong University of Science and Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Remote sensing image progressive transmission based on socket with retry broken downloads

Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE (2014)

Haicheng Qu Yu Meng Wanjun Liu Xiaochen Shan Jie Yu

In order to meet the different needs for different users to the quality of remote sensing images in heterogeneous network environments, an online remote sensing image progressive transmission model is constructed in which remote sensing image compression and decompression are synchronized with transmission. At the same time, a pipeline-based multi-threaded acceleration method has been proposed through solving the asynchronous problem between compression decompression and transmission to improve the efficiency of remote sensing progressive transmission. At last, an idea of retry broken downloads transmission interruption has been implemented to improve end-user interactive experience. Experimental results show that the whole processing speed has been improved nearly twice without reducing image transmission quality by using the proposed progressive transmission and real-time compression model.

10.1117/12.2057441

Cite

Citations (0)

DDFAV: Remote Sensing Large Vision Language Models Dataset and Evaluation Benchmark

arXiv (Cornell University) (2024)

Haodong Li Haicheng Qu Xiaofeng Zhang

With the rapid development of large vision language models (LVLMs), these models have shown excellent results in various multimodal tasks. Since LVLMs are prone to hallucinations and there are currently few datasets and evaluation methods specifically designed for remote sensing, their performance is typically poor when applied to remote sensing tasks. To address these issues, this paper introduces a high quality remote sensing LVLMs dataset, DDFAV, created using data augmentation and data mixing strategies. Next, a training instruction set is produced based on some high-quality remote sensing images selected from the proposed dataset. Finally, we develop a remote sensing LVLMs hallucination evaluation method RSPOPE based on the proposed dataset and evaluate the zero-shot capabilities of different LVLMs. Our proposed dataset, instruction set, and evaluation method files are available at https://github.com/HaodongLi2024/rspope.

Benchmark (surveying)

10.48550/arxiv.2411.02733

Cite

Citations (0)

Analysis and Extraction of a Characteristic Radius of the Trajectory Charging in a Blast Furnace Charging Process

Computer Engineering and Science (2010)

Haicheng Qu

As for the difficulty in analysing the trajectory charging of the current blast furnace,the pattern recognition technique is introduced and the mesurement method of the polar coordinate laser grid is proposed.To contrast the edge analysis of trajectory charging,we calculate the threshold value by linear probing.According to the image quality of trajectory charging in the difference spectrum,the characteristic radius is analysed and defined at the same time to different lengths of the characteristic radius for attaining accurate trajectory charging data information,the trajectory charging in polar coordinate is reconstructed,and the automatic control information of the blast furnace is obtanied.

Polar coordinate system

Source

Cite

Citations (0)

Semantic web service discovery algorithm and its application on the intelligent automotive manufacturing system

Guo Wen-yue Haicheng Qu Hong Chen

In order to solve the existing problems of many automobile manufacturing enterprises, there are three aspects of this problem have to be addressed. Firstly, the paper briefly introduces the research status and existing problems of semantic web service discovery, then the Service Profile of Ontology Web Language for Services (OWL-S)specification is extended. and a semantic web service discovery algorithm based on OWL-S is proposed, which takes service category matching, service functionality matching and quality of service matching into account. Afterwards, an intelligent automotive manufacturing system based on the semantic web service integration is designed and implemented. Furthermore, the proposed algorithm is applied to the system, and both the feasibility and effectiveness are experimentally verified. In a word, the results indicate that the system can intelligently execute service requests, matching and call, which ensures the inter-enterprise manufacturing information can be exchanged and shared in knowledge level.

Service discovery

10.1109/icime.2010.5478166

Cite

Citations (20)

Pruning Networks Using Filters Similarity Stability

Haicheng Qu Xuecong Zhang

Current filter pruning methods rely too much on pretrained weights and have many super parameters, resulting in obvious performance degradation and too long parameters adjustment time. In our research, we found that the cosine similarity distribution between filters can achieve stable in a few epochs during training. Therefore, a cluster pruning method named ECP(Early Cluster Pruning) based on the cosine similarity between filters in the early stage of training is proposed to compress the deep neural networks. First, in the early stage of training, the filters were clustered with a gradually increasing threshold, and then the reserved filters were selected randomly in each cluster. The pruned models could be obtained with only a few super parameters and a single training progress, leading to an obvious reduction in algorithmic complexity and large savings in training time. The experimental results on CIFAR-10 and CIFAR-100 datasets show that ECP method outperforms recent pruning methods in terms of model accuracy maintenance, training time, and model compression rate.

Pruning

Cosine similarity

Similarity (geometry)

Feature (linguistics)

10.1145/3584871.3584908

Cite

Citations (0)

Ships Detection in SAR Images Based on Anchor-Free Model With Mask Guidance Features

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (2021)

Haicheng Qu Lei Shen Wei Guo Junkai Wang

Ship targets in synthetic aperture radar (SAR) images have various scales. The detection model based on anchor boxes requires manual design of candidate boxes, which are fixed and cannot completely match all kinds of targets. Instead, large of anchor boxes with different sizes also result in large amounts of computing resources being consumed. Another potential issue comes from complex background information of near-coast scenes, which leads to ship targets being unrecognized because the background contains similar appearing objects. Therefore, this article proposes an anchor-free detection model based on mask guidance features, which achieves detection mainly through three modifications. First, feature maps of multiple scales are fused to obtain high-resolution feature maps containing rich semantic information. Second, a transformer encoder module is introduced to focus on the context relationship between the target object and the global image and to enhance the dependence between ship targets. Third, the mask guide feature is used to highlight the positions of the target in the feature map, and a loss function in the mask guide mechanism is designed to optimize the mask feature map to reduce false detections and missed detections. Testing the model on the public dataset SAR ship detection dataset, the model's detection accuracy reached 96.17%, with its accuracy on small-size ships reaching 96.11% and 97.84% on large ships.

Feature (linguistics)

10.1109/jstars.2021.3137390

Cite

Citations (17)

Multi-scale cost volumes information sharing based multi-view stereo reconstructed model

Journal of Image and Graphics (2022)

Wanjun Liu Junkai Wang Haicheng Qu

10.11834/jig.210774

Cite

Citations (2)

Hyperspectral image classification based on dominant sets clustering and Markov random fields

Guotu ziyuan yaogan (2019)

Haicheng Qu Yue Guo Yuanyuan Wang

Markov random field

10.6046/gtzyyg.2019.02.04

Cite

Citations (0)

Metric Learning with Neural Network for Modbus/TCP Anomaly Detection

Haicheng Qu Jianzhong Zhou Jitao Qin

In cyber security field, anomaly detection is triggered when detected network data traffic behaves obviously differently from normal data traffic. Traditional approaches typically create or define the normal pattern for the data and compare the normal pattern with the detected object. When a significantly different object appears, it is regarded as abnormal data. This paper proposes a novel neural network structure for anomaly detection, called a metric learning network, which aims to directly learn the differences between abnormal and normal data rather than set up a normal pattern. The network comprises an auto-encoder, which is used to encode the abnormal and normal data, and a metric learning component, which is designed to understand the difference between abnormal and normal data via a comparison approach. A deviation score is produced by the metric learning component to recognize the detected object. Research based on the Modbus/Transmission Control Protocol (TCP) network demonstrates that this approach can not only learn the difference between normal data and outliers, but is suitable for anomaly detection tasks. Our method has greater overall detection rates than a baseline model.

Modbus

10.1145/3411016.3411160

Cite

Citations (1)

劣化関数とモルフォロジーフィルタリングの画像除去アルゴリズム【JST・京大機械翻訳】

Zhongguo Tuxiang Tuxing Xuebao (2016)

Wanjun Liu Zhao Qingguo Haicheng Qu

Source

Cite

Citations (0)