Sketch-QNet: A Quadruplet ConvNet for Color Sketch-Based Image Retrieval

Abstract

Architectures based on siamese networks with triplet loss have shown outstanding performance on the image-based similarity search problem. This approach attempts to discriminate between positive (relevant) and negative (irrelevant) items. However, it undergoes a critical weakness. Given a query, it cannot discriminate weakly relevant items, for instance, items of the same type but different color or texture as the given query, which could be a serious limitation for many real-world search applications. Therefore, in this work, we present a quadruplet-based architecture that overcomes the aforementioned weakness. Moreover, we present an instance of this quadruplet network, which we call Sketch-QNet, to deal with the color sketch-based image retrieval (CSBIR) problem, achieving new state-of- the-art results.

Cite

Text

Fuentes and Saavedra. "Sketch-QNet: A Quadruplet ConvNet for Color Sketch-Based Image Retrieval." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021. doi:10.1109/CVPRW53098.2021.00242

Markdown

[Fuentes and Saavedra. "Sketch-QNet: A Quadruplet ConvNet for Color Sketch-Based Image Retrieval." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021.](https://mlanthology.org/cvprw/2021/fuentes2021cvprw-sketchqnet/) doi:10.1109/CVPRW53098.2021.00242

BibTeX

@inproceedings{fuentes2021cvprw-sketchqnet,
  title     = {{Sketch-QNet: A Quadruplet ConvNet for Color Sketch-Based Image Retrieval}},
  author    = {Fuentes, Anibal and Saavedra, José M.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2021},
  pages     = {2134-2141},
  doi       = {10.1109/CVPRW53098.2021.00242},
  url       = {https://mlanthology.org/cvprw/2021/fuentes2021cvprw-sketchqnet/}
}