Improvement of spatio-temporal prediction model
DOI:
https://doi.org/10.54097/ednk5k86Keywords:
Spatiotemporal prediction, Transformer, Cross-attention, Deep Learning, Feature alignment.Abstract
With the advancement of big data technology, spatiotemporal prediction models are playing an increasingly significant role in areas such as traffic flow forecasting, weather monitoring, and urban management. However, traditional spatiotemporal prediction models still face limitations in handling high-dimensional data, capturing nonlinear relationships, and considering long-term dependencies. This paper proposes an improved spatiotemporal prediction framework by integrating the strengths and weaknesses of existing models, such as ARIMA, LSTM, and GCN. Experimental results show that this method outperforms mainstream models on multiple real datasets, demonstrating higher prediction accuracy and stronger generalization capabilities. This research provides new insights for further optimizing spatiotemporal prediction models and offers robust support for practical applications.
Downloads
References
[1] Meng Weijun, Shan Lianlei, Ma Sugang, Liu Dan, Hu Bin. DLNet: A Dual-Level Network with Self-and Cross-Attention for High-Resolution Remote Sensing Segmentation. Remote Sensing, 2025, 17 (7): 1119.
[2] Zhou Xirui, Shan Lianlei, Gui Xiaolin. DynRsl-VLM: Enhancing Autonomous Driving Perception with Dynamic Resolution Vision-Language Models. arXiv preprint arXiv:2503.11265, 2025.
[3] Pi Ruochen, Shan Lianlei. Synthetic Lung X-ray Generation through Cross-Attention and Affinity Transformation. arXiv preprint arXiv:2503.07209, 2025.
[4] Du Bingyun, Shan Lianlei, Shao Xiaoyu, Zhang Dongyou, Wang Xinrui, Wu Jiaxi. Transform Dual-Branch Attention Net: Efficient Semantic Segmentation of Ultra-High-Resolution Remote Sensing Images. Remote Sensing, 2025, 17 (3): 540.
[5] Ji Yuyang, Shan Lianlei. LDNET: Semantic Segmentation of High-Resolution Images Via Learnable Patch Proposal and Dynamic Refinement. In 2024 IEEE International Conference on Multimedia and Expo (ICME), IEEE, 2024, pp. 1-6,
[6] Shan Lianlei, Zhou Wenzhang, Li Wei, Ding Xingyu. Organizing Background to Explore Latent Classes for Incremental Few-shot Semantic Segmentation. arXiv preprint arXiv:2405.19568, 2024.
[7] Shan Lianlei, Zhou Wenzhang, Li Wei, Ding Xingyu. Lifelong Learning and Selective Forgetting via Contrastive Strategy. arXiv preprint arXiv:2405.18663, 2024.
[8] Zhao Leo Shan Wenzhang Zhou Grace. Boosting General Trimap-free Matting in the Real-World Image. arXiv preprint arXiv:2405.17916, 2024.
[9] Shan Lianlei, Wang Weiqiang, Lv Ke, Luo Bin. Edge-guided and Class-balanced Active Learning for Semantic Segmentation of Aerial Images. arXiv preprint arXiv:2405.18078, 2024.
[10] Ding Xingyu, Shan Lianlei, Zhao Guiqin, Wu Meiqi, Zhou Wenzhang, Li Wei. The Binary Quantized Neural Network for Dense Prediction via Specially Designed Upsampling and Attention. arXiv preprint arXiv:2405.17776, 2024.
[11] Wu Weijia, Zhao Yuzhong, Li Zhuang, Shan Lianlei, Zhou Hong, Shou Mike Zhang. Continual learning for image segmentation with dynamic query. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 34 (6): 4874-4886.
[12] Shan Leo, Zhou Wenzhang, Zhao Grace. Incremental few shot semantic segmentation via class-agnostic mask proposal and language-driven classifier. In Proceedings of the 31st ACM International Conference on Multimedia, 2023, pp. 8561-8570.
[13] Shan Lianlei, Zhao Guiqin, Xie Jun, Cheng Peirui, Li Xiaobin, Wang Zhepeng. A Data-Related Patch Proposal for Semantic Segmentation of Aerial Images. IEEE Geoscience and Remote Sensing Letters, 2023, 20: 1-5.
[14] Zhao Guiqin, Shan Lianlei, Wang Weiqiang. End-to-End Remote Sensing Change Detection of Unregistered Bi-temporal Images for Natural Disasters. In International Conference on Artificial Neural Networks, 2023, pp. 259-270. Cham: Springer Nature Switzerland.
[15] Shan Lianlei, Wang Weiqiang, Lv Ke, Luo Bin. Boosting semantic segmentation of aerial images via decoupled and multilevel compaction and dispersion. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-16.
[16] Shan Lianlei, Wang Weiqiang, Lv Ke, Luo Bin. Class-incremental semantic segmentation of aerial images via pixel-level feature generation and task-wise distillation. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-17.
[17] Shan Lianlei, Wang Weiqiang. Mbnet: A multi-resolution branch network for semantic segmentation of ultra-high resolution images. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2022, pp. 2589-2593.
[18] Shan Lianlei, Wang Weiqiang, Lv Ke, Luo Bin. Class-incremental learning for semantic segmentation in aerial imagery via distillation in all aspects. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-12.
[19] Li Xiaobin, Shan Lianlei, Wang Weiqiang. Fusing multitask models by recursive least squares. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2021, pp. 3640-3644.
[20] Shan Lianlei, Li Xiaobin, Wang Weiqiang. Decouple the high-frequency and low-frequency information of images for semantic segmentation. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2021, pp. 1805-1809.
[21] Shan Lianlei, Wang Weiqiang. DenseNet-based land cover classification network with deep fusion. IEEE Geoscience and Remote Sensing Letters, 2021, 19: 1-5.
[22] Li Minglong, Shan Lianlei, Li Xiaobin, Bai Yang, Zhou Dengji, Wang Weiqiang, Lv Ke, Luo Bin, Chen Si-Bao. Global-local attention network for semantic segmentation in aerial images. In 2020 25th International Conference on Pattern Recognition (ICPR), IEEE, 2021, pp. 5704-5711.
[23] Li Xiaobin, Shan Lianlei, Li Minglong, Wang Weiqiang. Energy Minimum Regularization in Continual Learning. In 2020 25th International Conference on Pattern Recognition (ICPR), IEEE, 2021, pp. 6404-6409.
[24] Shan Lianlei, Li Minglong, Li Xiaobin, Bai Yang, Lv Ke, Luo Bin, Chen Si-Bao, Wang Weiqiang. Uhrsnet: A semantic segmentation network specifically for ultra-high-resolution images. In 2020 25th International Conference on Pattern Recognition (ICPR), IEEE, 2021, pp. 1460-1466.
[25] Yi-Ge, Ellen, and Leo Shawn. FlexDataset: Crafting Annotated Dataset Generation for Diverse Applications. Proceedings of the AAAI Conference on Artificial Intelligence, 2025, 39 (9).
[26] Li Minglong, Shan Lianlei, Wang Weiqiang, Ke Lv, Bin Luo, Si-Bao Chen. Building Lightweight Semantic Segmentation Models for Aerial Images Using Dual Relation Distillation. arXiv preprint arXiv:2506.20688, 2025.
[27] Li Minglong, Shan Lianlei, Wang Weiqiang, Lv Ke, Luo Bin, Chen Si-Bao. Building Lightweight Semantic Segmentation Models for Aerial Images Using Dual Relation Distillation. arXiv preprint arXiv:2506.20688, 2025.
[28] Xie Yu, Ren Xingkai, Qi Ying, Hu Yao, Shan Lianlei. RecLLM-R1: A Two-Stage Training Paradigm with Reinforcement Learning and Chain-of-Thought v1. arXiv preprint arXiv:2506.19235, 2025.
[29] Chen Yi, Shan Lianlei. A Global-Local Cross-Attention Network for Ultra-high Resolution Remote Sensing Image Semantic Segmentation. arXiv preprint arXiv:2506.19406, 2025.
[30] Chen Hengzhi, Feng Liqian, Wu Wenhua, Zhu Xiaogang, Leo Shawn, Hu Kun. F2Net: A Frequency-Fused Network for Ultra-High Resolution Remote Sensing Segmentation. arXiv preprint arXiv:2506.07847, 2025.
[31] Yi Qiang, Shan Lianlei. GeoLocSFT: Efficient Visual Geolocation via Supervised Fine-Tuning of Multimodal Foundation Models. arXiv preprint arXiv:2506.01277, 2025.
[32] Sun Chengsong, Li Weiping, Li Xiang, Liu YuanKun, Shan Lianlei. Gmm-based comprehensive feature extraction and relative distance preservation for few-shot cross-modal retrieval. arXiv preprint arXiv:2505.13306, 2025.
[33] Luo Hailong, Wu Bin, Jia Hongyong, Zhu Qingqing, Shan Lianlei. LLM-CoT Enhanced Graph Neural Recommendation with Harmonized Group Policy Optimization. arXiv preprint arXiv:2505.12396, 2025.
[34] Liu Shiyu, Lianlei Shan. NeuroVoxel-LM: Language-Aligned 3D Perception via Dynamic Voxelization and Meta-Embedding. arXiv preprint arXiv:2507.20110, 2025.
[35] Zhang Jinbo, Chen Min, Zhao Yitao, Shan Lianlei, Li Caiyi, Hu Han. Asymmetric Mamba-CNN Collaborative Architecture for Large-Size Remote Sensing Image Semantic Segmentation. IEEE Transactions on Geoscience and Remote Sensing, 2025.
[36] Wang, Xingjun, Lianlei Shan. GDGS: 3D Gaussian Splatting Via Geometry-Guided Initialization And Dynamic Density Control. arXiv preprint arXiv:2507.00363, 2025.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Highlights in Science, Engineering and Technology
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
