Revisiting Low-Resolution Images Retrieval with Attention Mechanism and Contrastive Learning

Round 1

Reviewer 1 Report

Personally, I found the paper well-written and engaging. The comments I have are minor:

1. The programming language and versions of the various components  (e.g. the Visual Transformer) of this tool should be specified.
2. It would be clearer if a network diagram could be included
3. Could the authors could make the source code available as open source?

Author Response

Thank you for your precious comments on the manuscript entitled "Revisiting low-resolution images retrieval with attention mechanism and contrastive learning". We are pleased to answer those reviews as below.

Author Response File: Author Response.docx

Reviewer 2 Report

1. This work tries to find a solution for the challenging problem of category image retrieval tasks on low-resolution images   ---Can you mention 1/2 real life application/scenario??
2.  It seems the only major contribution is to  use a powerful pre-trained encoder to extract visual representations and fine-tune contrastive learning to learn embeddings for feature matching.  If you have more contribution then can you  please put as bullet points at the end of Introduction section.
3. Can you explain further Fig. 1, Not clear about the failure cases here.
4. I like the idea of using all recent CNN models...you may expand or describe a bit  with few lines each model.
5. For a realistic retrieval application , it is not a good idea to simply use an exhaustive search using ?2 similarity.  What is the size of the final feature vector. Can you use some kind of indexing mechanism (such, as tree structure) for faster r retrieval?

Author Response

Thank you for your precious comments on the manuscript entitled "Revisiting low-resolution images retrieval with attention mechanism and contrastive learning". We are pleased to answer those reviews as below.

Author Response File: Author Response.docx