[TOC]

  1. Title: Semantic-Aligned Fusion Transformer for One Shot Object Detection
  2. Author: Yizhou Zhao et. al.
  3. Publish Year: 2022
  4. Review Date: Mon, Oct 24, 2022

https://arxiv.org/pdf/2203.09093v2.pdf

Summary of paper

Motivation

  • with extreme data scarcity, current approaches, explore various feature fusions to obtain directly transferable meta-knowledge
  • in this paper, they, attribute the previous limitation to inappropriate correlation methods that misalign query-support semantics by overlooking spatial structure and scale variances.

image-20221025164925086

image-20221025165002340