Task-Driven Learned Hyperspectral Data Reduction Using End-to-End Supervised Deep Learning

被引:8
|
作者
Zeegers, Mathe T. [1 ]
Pelt, Daniel M. [1 ,2 ]
van Leeuwen, Tristan [1 ,3 ]
van Liere, Robert [1 ,4 ]
Batenburg, Kees Joost [1 ,2 ]
机构
[1] Ctr Wiskunde & Informat, Sci Pk 123, NL-1098 XG Amsterdam, Netherlands
[2] Leiden Inst Adv Comp Sci, Niels Bohrweg 1, NL-2333 CA Leiden, Netherlands
[3] Univ Utrecht, Math Inst, Budapestlaan 6, NL-3584 CD Utrecht, Netherlands
[4] Tech Univ Eindhoven, Fac Wiskunde & Informat, Groene Loper 5, NL-5612 AZ Eindhoven, Netherlands
关键词
hyperspectral imaging; feature extraction; compression; machine learning; deep learning; convolutional neural network; segmentation; IMAGE CLASSIFICATION; FEATURE-EXTRACTION; FUSION;
D O I
10.3390/jimaging6120132
中图分类号
TB8 [摄影技术];
学科分类号
0804 ;
摘要
An important challenge in hyperspectral imaging tasks is to cope with the large number of spectral bins. Common spectral data reduction methods do not take prior knowledge about the task into account. Consequently, sparsely occurring features that may be essential for the imaging task may not be preserved in the data reduction step. Convolutional neural network (CNN) approaches are capable of learning the specific features relevant to the particular imaging task, but applying them directly to the spectral input data is constrained by the computational efficiency. We propose a novel supervised deep learning approach for combining data reduction and image analysis in an end-to-end architecture. In our approach, the neural network component that performs the reduction is trained such that image features most relevant for the task are preserved in the reduction step. Results for two convolutional neural network architectures and two types of generated datasets show that the proposed Data Reduction CNN (DRCNN) approach can produce more accurate results than existing popular data reduction methods, and can be used in a wide range of problem settings. The integration of knowledge about the task allows for more image compression and higher accuracies compared to standard data reduction methods.
引用
收藏
页数:28
相关论文
共 50 条
  • [1] TDMF: TASK-DRIVEN MULTILEVEL FRAMEWORK FOR END-TO-END SPEAKER VERIFICATION
    Chen, Chen
    Han, Jiqing
    2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2020, : 6809 - 6813
  • [2] End-to-End Communication With Task-Driven CSI Acquisition in Multiple-Antenna Systems
    Zhang, Jinya
    Guo, Jiajia
    Wen, Chao-Kai
    Jin, Shi
    IEEE WIRELESS COMMUNICATIONS LETTERS, 2024, 13 (10) : 2897 - 2901
  • [3] End-to-End Deep Reinforcement Learning based Recommendation with Supervised Embedding
    Liu, Feng
    Guo, Huifeng
    Li, Xutao
    Tang, Ruiming
    Ye, Yunming
    He, Xiuqiang
    PROCEEDINGS OF THE 13TH INTERNATIONAL CONFERENCE ON WEB SEARCH AND DATA MINING (WSDM '20), 2020, : 384 - 392
  • [4] Deep-learning-driven end-to-end metalens imaging
    Joonhyuk Seo
    Jaegang Jo
    Joohoon Kim
    Joonho Kang
    Chanik Kang
    SeongWon Moon
    Eunji Lee
    Jehyeong Hong
    Junsuk Rho
    Haejun Chung
    Advanced Photonics, 2024, 6 (06) : 71 - 83
  • [5] Deep-learning-driven end-to-end metalens imaging
    Seo, Joonhyuk
    Jo, Jaegang
    Kim, Joohoon
    Kang, Joonho
    Kang, Chanik
    Moon, Seong-Won
    Lee, Eunji
    Hong, Jehyeong
    Rho, Junsuk
    Chung, Haejun
    ADVANCED PHOTONICS, 2024, 6 (06):
  • [6] Semi-Supervised Learning with Data Augmentation for End-to-End ASR
    Weninger, Felix
    Mana, Franco
    Gemello, Roberto
    Andres-Ferrer, Jesus
    Zhan, Puming
    INTERSPEECH 2020, 2020, : 2802 - 2806
  • [7] Automated Classification Using End-to-End Deep Learning
    Jaipurkar, Shobhit Sandeep
    Jie, Wang
    Zeng, Zeng
    Gee, Teo Sin
    Veeravalli, Bharadwaj
    Chua, Matthew
    2018 40TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY (EMBC), 2018, : 706 - 709
  • [8] SCL-Net: An End-to-End Supervised Contrastive Learning Network for Hyperspectral Image Classification
    Lu, Ting
    Hu, Yaochen
    Fu, Wei
    Ding, Kexin
    Bai, Beifang
    Fang, Leyuan
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2022, 60
  • [9] End-to-end deep learning pipeline for prediction of IVF outcomes using ultrasound data
    Vali, M.
    Yang, J.
    Vali, S.
    Azevedo, T.
    Lio, P.
    Thum, Y.
    BJOG-AN INTERNATIONAL JOURNAL OF OBSTETRICS AND GYNAECOLOGY, 2024, 131 : 187 - 187
  • [10] Traffic Signal Recognition Using End-to-End Deep Learning
    Sarker, Tonmoy
    Meng, Xiangyu
    TRAN-SET 2022, 2022, : 182 - 191