大模型赋能智能摄影测量：现状、挑战与前景

doi:10.11947/j.AGCS.2024.20240068.

摘要/Abstract

摘要：

大模型从深度学习和迁移学习技术发展而来，依靠大量的训练数据和庞大的参数容量产生规模效应，从而激发了模型的涌现能力，在众多下游任务中展现了强大的泛化性和适应性。以ChatGPT、SAM为代表的大模型标志着通用人工智能时代的到来，为地球空间信息处理的自动化与智能化提供了新的理论与技术。为了进一步探索大模型赋能泛摄影测量领域的方法与途径，本文回顾了摄影测量领域的基本问题和任务内涵，总结了深度学习方法在摄影测量智能处理中的研究成果，分析了面向特定任务的监督预训练方法的优势与局限；阐述了通用人工智能大模型的特点及研究进展，关注大模型在基础视觉任务中的场景泛化性以及三维表征方面的潜力；从训练数据、模型微调策略和异构多模态数据融合处理3个方面，探讨了大模型技术在摄影测量领域当前面临的挑战与发展趋势。

关键词: 大模型, 智能摄影测量, 深度学习, 多模态

Abstract:

Developed from deep learning and transfer learning techniques, large models leverage vast training datasets and immense parameter capacities to create scale effects, thus inspiring the model's emergent capabilities and demonstrating strong generalization and adaptability in numerous downstream tasks. Large models, represented by ChatGPT and SAM, signify the arrival of the era of general artificial intelligence, providing new theories and techniques for the automation and intelligence of Earth's spatial information processing. To further explore the methods and pathways for large models to empower the field of photogrammetry, this paper reviews the basic problems and mission tasks in the field of photogrammetry, summarizes the research achievements of deep learning methods in intelligent photogrammetric processing, analyzes the advantages and limitations of supervised pre-training methods aimed at specific tasks; Besides, we elaborates on the characteristics and research progress of general artificial intelligence large models, focusing on the generalizability of large models in basic visual tasks and the potential in three-dimensional representation; Finally, this paper explores the current challenges and future trends of large model technologies in the field of photogrammetry, from the perspectives of training data, model fine-tuning strategies, and heterogeneous multi-modal data fusion strategies.

Key words: large models, intelligent photogrammetry, deep learning, multi-modal

中图分类号:

P237

王密, 程昫, 潘俊, 皮英冬, 肖晶. 大模型赋能智能摄影测量：现状、挑战与前景[J]. 测绘学报, 2024, 53(10): 1955-1966.

Mi WANG, Xu CHENG, Jun PAN, Yingdong PI, Jing XIAO. Large models enabling intelligent photogrammetry: status, challenges and prospects[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(10): 1955-1966.

图/表 2

参考文献 81

[1]	李德仁. 摄影测量与遥感的现状及发展趋势[J]. 武汉测绘科技大学学报, 2000, 25(1):1-6.
	LI Deren. Towards photogrammetry and remote sensing: status and future development[J]. Journal of Wuhan Technical University of Surveying and Mapping, 2000, 25(1):1-6.
[2]	宁津生, 杨凯. 从数字化测绘到信息化测绘的测绘学科新进展[J]. 测绘科学, 2007, 32(2):5-11.
	NING Jinsheng, YANG Kai. The newest progress of surveying & mapping discipline from digital style to informatization stage[J]. Science of Surveying and Mapping, 2007, 32(2):5-11.
[3]	李德仁. 从测绘学到地球空间信息智能服务科学[J]. 测绘学报, 2017, 46(10):1207-1212. DOI:. doi: 10.11947/j.AGCS.2017.20170263
	LI Deren. From geomatics to geospatial intelligent service science[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(10):1207-1212. DOI:. doi: 10.11947/j.AGCS.2017.20170263
[4]	张永军, 万一, 史文中, 等. 多源卫星影像的摄影测量遥感智能处理技术框架与初步实践[J]. 测绘学报, 2021, 50(8):1068-1083. DOI:. doi: 10.11947/j.AGCS.2021.20210079
	ZHANG Yongjun, WAN Yi, SHI Wenzhong, et al. Technical framework and preliminary practices of photogrammetric remote sensing intelligent processing of multi-source satellite images[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(8):1068-1083. DOI:. doi: 10.11947/j.AGCS.2021.20210079
[5]	DENG Jia, DONG Wei, SOCHER R, et al. ImageNet: a large-scale hierarchical image database[C]//Proceedings of 2009 IEEE Confe-rence on Computer Vision and Pattern Recognition. Miami: IEEE, 2009: 248-255.
[6]	BROWN T, MANN B, RYDER N, et al. Language models are few-shot learners[J]. Advances in neural information processing systems, 2020, 33:1877-1901.
[7]	BOMMASANI R, HUDSON D A, ADELI E, et al. On the opportunities and risks of foundation models[EB/OL]. [2023-11-28]. http://arxiv.org/abs/2108.07258.
[8]	龚健雅, 季顺平. 摄影测量与深度学习[J]. 测绘学报, 2018, 47(6):693-704. DOI:. doi: 10.11947/j.AGCS.2018.20170640
	GONG Jianya, JI Shunping. Photogrammetry and deep learning[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(6):693-704. DOI:. doi: 10.11947/j.AGCS.2018.20170640
[9]	张良培, 张乐飞, 袁强强. 遥感大模型：进展与前瞻[J]. 武汉大学学报(信息科学版), 2023, 48(10):1574-1581.
	ZHANG Liangpei, ZHANG Lefei, YUAN Qiangqiang. Large remote sensing model: progress and prospects[J]. Geomatics and Information Science of Wuhan University, 2023, 48(10):1574-1581.
[10]	杨必胜, 陈一平, 邹勤. 从大模型看测绘时空信息智能处理的机遇和挑战[J]. 武汉大学学报(信息科学版), 2023, 48(11):1756-1768.
	YANG Bisheng, CHEN Yiping, ZOU Qin. Opportunities and challenges of spatiotemporal information intelligent processing of surveying and mapping in the era of large models[J]. Geomatics and Information Science of Wuhan University, 2023, 48(11):1756-1768.
[11]	DETONE D, MALISIEWICZ T, RABINOVICH A. SuperPoint: self-supervised interest point detection and description[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE, 2018: 224-236.
[12]	SARLIN P E, DETONE D, MALISIEWICZ T, et al. SuperGlue: learning feature matching with graph neural networks[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 4938-4947.
[13]	赵伍迪, 李山山, 李安, 等. 结合深度学习的高光谱与多源遥感数据融合分类[J]. 遥感学报, 25(7):1489-1502.
	ZHAO Wudi, LI Shanshan, LI An, et al. Deep fusion of hyperspectral images and multi-source remote sensing data for classification with convolutional neural network[J]. National Remote Sensing Bulletin, 25(7):1489-1502.
[14]	YANG Y, JIAO L, LIU F, et al. An explainable spatial-frequency multi-scale transformer for remote sensing scene classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61:5907515.
[15]	张祖勋, 姜慧伟, 庞世燕, 等. 多时相遥感影像的变化检测研究现状与展望[J]. 测绘学报, 2022, 51(7):1091-1107. DOI:. doi: 10.11947/j.AGCS.2022.20220070
	ZHANG Zuxun, JIANG Huiwei, PANG Shiyan, et al. Review and prospect in change detection of multi-temporal remote sensing images[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(7):1091-1107. DOI:. doi: 10.11947/j.AGCS.2022.20220070
[16]	SU Hang, MAJI S, KALOGERAKIS E, et al. Multi-view convolutional neural networks for 3D shape recognition[C]//Proceedings of 2015 IEEE International Conference on Computer Vision. Santiago: IEEE, 2015: 945-953.
[17]	KALOGERAKIS E, AVERKIOU M, MAJI S, et al. 3D shape segmentation with projective convolutional networks[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 3779-3788.
[18]	WU Zhirong, SONG Shuran, KHOSLA A, et al. 3D ShapeNets: a deep representation for volumetric shapes[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 1912-1920.
[19]	BROCK A, LIM T, RITCHIE J, et al. Generative and discriminative voxel modeling with convolutional neural networks[EB/OL]. [2024-02-15]. https://arxiv.org/abs/1608.04236.
[20]	CHARLES R Q, HAO Su, MO Kaichun, et al. PointNet: deep learning on point sets for 3D classification and segmentation[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 652-660.
[21]	QI C R, YI Li, SU Hao, et al. PointNet++: deep hierarchical feature learning on point sets in a metric space[EB/OL]. [2024-02-15]. https://arxiv.org/abs/1706.02413v1.
[22]	HANOCKA R, HERTZ A, FISH N, et al. MeshCNN: a network with an edge[J]. ACM Transactions on Graphics, 2019, 38(4):1-12.
[23]	YAO Yao, LUO Zixin, LI Shiwei, et al. MVSNet: depth inference for unstructured multi-view stereo[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2018: 767-783.
[24]	GU Xiaodong, FAN Zhiwen, ZHU Siyu, et al. Cascade cost volume for high-resolution multi-view stereo and stereo matching[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. Seattle: IEEE, 2020: 2495-2504.
[25]	YANG Jiayu, MAO Wei, ALVAREZ J M, et al. Cost volume pyramid based depth inference for multi-view stereo[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 4877-4886.
[26]	LUO Keyang, GUAN Tao, JU Lili, et al. P-MVSNet: learning patch-wise matching confidence aggregation for multi-view stereo[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 10452-10461.
[27]	XU Qingshan, SU Wanjuan, QI Yuhang, et al. Learning inverse depth regression for pixelwise visibility-aware multi-view stereo networks[J]. International Journal of Computer Vision, 2022, 130(8):2040-2059.
[28]	ZHANG Jingyang, LI Shiwei, LUO Zixin, et al. Vis-MVSNet: visibility-aware multi-view stereo network[J]. International Journal of Computer Vision, 2023, 131(1):199-214.
[29]	GKIOXARI G, JOHNSON J, MALIK J. Mesh R-CNN[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 9785-9795.
[30]	EIGEN D, PUHRSCH C, FERGUS R. Depth map prediction from a single image using a multi-scale deep network[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems. Cambridge: MIT Press, 2014: 2366-2374.
[31]	CHOY C B, XU Danfei, GWAK J, et al. 3D-R2N2: a unified approach for single and multi-view 3D object reconstruction[M]//Lecture notes in computer science. Cham: Springer International Publishing, 2016: 628-644.
[32]	FAN Haoqiang, SU Hao, GUIBAS L. A point set generation network for 3D object reconstruction from a single image[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 605-613.
[33]	WANG Nanyang, ZHANG Yinda, LI Zhuwen, et al. Pixel2Mesh: generating 3D mesh models from single RGB images[M]//Lecture notes in computer science. Cham: Springer International Publishing, 2018: 55-71.
[34]	RADFORD A, KIM J W, HALLACY C, et al. Learning transferable visual models from natural language supervision[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2103.00020.
[35]	ZHAI Xiaohua, WANG Xiao, MUSTAFA B, et al. LiT: zero-shot transfer with locked-image text tuning[C]//Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 18123-18133.
[36]	TSIMPOUKELLI M, MENICK J L, CABI S, et al. Multimodal few-shot learning with frozen language models[J]. Advances in Neural Information Processing Systems, 2021, 34:200-212.
[37]	ALAYRAC J B, DONAHUE J, LUC P, et al. Flamingo: a visual language model for few-shot learning[J]. Advances in Neural Information Processing Systems, 2022, 35:23716-23736.
[38]	YAO Lewei, HUANG Runhu, HOU Lu, et al. FILIP: fine-grained interactive language-image pre-training[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2111.07783.
[39]	WANG X, ZHANG X, CAO Y, et al. SegGPT: segmenting everything in context[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2304.03284.
[40]	KIRILLOV A, MINTUN E, RAVI N, et al. Segment anything[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2304.02643v1.
[41]	LYU Chenyang, WU Minghao, WANG Longyue, et al. Macaw-LLM: multi-modal language modeling with image, audio, video, and text integration[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2306.09093v1.
[42]	GIRDHAR R, EL-NOUBY A, LIU Zhuang, et al. ImageBind one embedding space to bind them all[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 15180-15190.
[43]	ZHANG Yiyuan, GONG Kaixiong, ZHANG Kaipeng, et al. Meta-transformer: a unified framework for multimodal learning[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2307.10802v1.
[44]	XUE Le, GAO Mingfei, XING Chen, et al. ULIP: learning a unified representation of language, images, and point clouds for 3D understanding[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 1179-1189.
[45]	ZHOU J, WANG J, MA B, et al. Uni3D: exploring unified 3D representation at scale[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2310.06773.
[46]	SHORTEN C, KHOSHGOFTAAR T M. A survey on image data augmentation for deep learning[J]. Journal of Big Data, 2019, 6(1):60.
[47]	TRABUCCO B, DOHERTY K, GURINAS M, et al. Effective data augmentation with diffusion models[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2302.07944v2.
[48]	SHUMAILOV I, SHUMAYLOV Z, ZHAO Yiren, et al. The curse of recursion: training on generated data makes models forget[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2305.17493v3.
[49]	SUN Xian, WANG Peijin, LU Wanxuan, et al. RingMo: a remote sensing foundation model with masked image modeling[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61:5612822.
[50]	JIA Chao, YANG Yinfei, XIA Ye, et al. Scaling up visual and vision-language representation learning with noisy text supervision[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2102.05918.
[51]	XIANG Shouxing, YE Xi, XIA Jiazhi, et al. Interactive correction of mislabeled training data[C]//Proceedings of 2019 IEEE Confe-rence on Visual Analytics Science and Technology. Vancouver: IEEE, 2019: 57-68.
[52]	REIF E, KAHNG M, PETRIDIS S. Visualizing linguistic diversity of text datasets synthesized by large language models[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2305.11364v2.
[53]	KOLVE E, MOTTAGHI R, HAN W, et al. AI2-THOR: an interactive 3D environment for visual AI[EB/OL]. [2024-02-15]. https://arxiv.org/abs/1712.05474v4.
[54]	LI Chengshu, XIA Fei, MARTÍN-MARTÍN R, et al. iGibson 2.0: object-centric simulation for robot learning of everyday household tasks[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2108.03272v4.
[55]	RAMRAKHYA R, UNDERSANDER E, BATRA D, et al. Habitat-web: learning embodied object-search strategies from human demonstrations at scale[C]//Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 5173-5183.
[56]	SZOT A, CLEGG A, UNDERSANDER E, et al. Habitat 2.0: training home assistants to rearrange their habitat[J]. Advances in Neural Information Processing Systems, 2021, 34:251-266.
[57]	DEITKE M, SCHWENK D, SALVADOR J, et al. Objaverse: a universe of annotated 3D objects[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 13142-13153.
[58]	LING Lu, SHENG Yichen, TU Zhi, et al. DL3DV-10K: a large-scale scene dataset for deep learning-based 3D vision[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2312.16256v2.
[59]	YANG Y, WU X, HE T, et al. SAM3D: segment anything in 3D scenes[EB/OL]. [2023-11-27]. http://arxiv.org/abs/2306.03908.
[60]	CEN J, ZHOU Z, FANG J, et al. Segment anything in 3D with NeRFs[EB/OL]. [2023-11-27]. http://arxiv.org/abs/2304.12308.
[61]	MARI R, FACCIOLO G, EHRET T. Sat-NeRF: learning multi-view satellite photogrammetry with transient objects and shadow modeling using RPC cameras[C]//Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. New Orleans: IEEE, 2022: 1311-1321.
[62]	XU Linning, XIANGLI Yuanbo, PENG Sida, et al. Grid-guided neural radiance fields for large urban scenes[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023.
[63]	TANCIK M, CASSER V, YAN Xinchen, et al. Block-NeRF: scalable large scene neural view synthesis[C]//2022 IEEE/CVF Confe-rence on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 8248-8258.
[64]	LIALIN V, DESHPANDE V, RUMSHISKY A. Scaling down to scale up: a guide to parameter-efficient fine-tuning[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2303.15647v1.
[65]	HU J E, SHEN Yelong, WALLIS P, et al. LoRA: low-rank adaptation of large language models[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2106.09685.
[66]	HOULSBY N, GIURGIU A, JASTRZEBSKI S, et al. Parameter-efficient transfer learning for NLP[EB/OL]. [2024-02-15]. https://arxiv.org/abs/1902.00751.
[67]	PFEIFFER J, RÜCKLÉ A, POTH C, et al. AdapterHub: a framework for adapting transformers[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2007.07779v3.
[68]	WANG Xinlong, WANG Wen, CAO Yue, et al. Images speak in images: a generalist painter for in-context visual learning[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 6830-6839.
[69]	SAITO K, SOHN K, ZHANG Xiang, et al. Prefix conditioning unifies language and label supervision[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 2861-2870.
[70]	CHEN Keyan, LIU Chenyang, CHEN Hao, et al. RSPrompter: learning to prompt for remote sensing instance segmentation based on visual foundation model[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2306.16269v2.
[71]	WEI J, WANG Xuezhi, SCHUURMANS D, et al. Chain-of-thought prompting elicits reasoning in large language models[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2201.11903v6.
[72]	OUYANG L, WU J, JIANG X, et al. Training language models to follow instructions with human feedback[J]. Advances in Neural Information Processing Systems, 2022, 35:27730-27744.
[73]	GAN Zhe, LI Linjie, LI Chunyuan, et al. Vision-language pre-training: basics, recent advances, and future trends[J]. Foundations and Trends® in Computer Graphics and Vision, 2022, 14(3/4):163-352.
[74]	SHEN Qiuhong, YANG Xingyi, WANG Xinchao. Anything-3D: towards single-view anything reconstruction in the wild[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2304.10261v1.
[75]	YANG X, ZHOU D, LIU S, et al. Deep model reassembly[J]. Advances in Neural Information Processing Systems, 2022, 35:25739-25753.
[76]	XI Zhiheng, CHEN Wenxiang, GUO Xin, et al. The rise and potential of large language model based agents: a survey[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2309.07864v3.
[77]	CONG Y, KHANNA S, MENG C, et al. Satmae: pre-training transformers for temporal and multi-spectral satellite imagery[J]. Advances in Neural Information Processing Systems, 2022, 35:197-211.
[78]	WANG Di, ZHANG Qiming, XU Yufei, et al. Advancing plain vision transformer toward remote sensing foundation model[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61:5607315.
[79]	LIU Fan, CHEN Delong, GUAN Z, et al. RemoteCLIP: a vision language foundation model for remote sensing[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2306.11029v4.
[80]	ZHANG Zilun, ZHAO Tiancheng, GUO Yulong, et al. RS5M: a large scale vision-language dataset for remote sensing vision-language foundation model[EB/OL]. [2024-02-15]. http://export.arxiv.org/abs/2306.11300v4.
[81]	WANG Di, ZHANG Jing, DU Bo, et al. SAMRS: scaling-up remote sensing segmentation dataset with segment anything model[EB/OL]. [2024-02-15]. https://arxiv.org/abs/2305.02034.