Target Detection Adapting to Spectral Variability in Multi-Temporal Hyperspectral Images Using Implicit Contrastive Learning
, Kun Gao 1,*, Junwei Wang 1, Pengyu Wang 1, Zibo Hu 1, Zhijia Yang 1, Xiaobin Zhao 2 and Wei Li 2Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis manuscript proposes the Implicit Contrastive Learning Module (ICLM) and the Local Spectral Similarity Constraint Loss (LSSC) to implement an Implicit Contrastive Learning-based Target Detector (ICLTD). Detailed experiments and derivations compellingly illustrate the effectiveness of the proposed method. Nonetheless, a few minor issues merit attention:
1. As an improvement of BN, whether ICLM can play a role in other HSI detection methods. Can it be used as a plug-and-play method and be widely used in other DL-based HTD methods?
2. In Figure 12, compared with GT, we notice that the proposed method LSSC also enhances some noise (A1, B2, U1). Is this an inherent flaw in the approach?
3. Given the relatively innovative nature of Implicit Contrastive Learning, could you provide a brief prospect to foster the development of Implicit Contrastive Learning in the HTD domain?
4. The slashes in the headers of Tables 6,7 and 8 may need to be corrected.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsSee attached file.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors answered all my questions.
