English

One-Shot Texture Retrieval with Global Context Metric

Computer Vision and Pattern Recognition 2020-04-14 v2 Artificial Intelligence

Abstract

In this paper, we tackle one-shot texture retrieval: given an example of a new reference texture, detect and segment all the pixels of the same texture category within an arbitrary image. To address this problem, we present an OS-TR network to encode both reference and query image, leading to achieve texture segmentation towards the reference category. Unlike the existing texture encoding methods that integrate CNN with orderless pooling, we propose a directionality-aware module to capture the texture variations at each direction, resulting in spatially invariant representation. To segment new categories given only few examples, we incorporate a self-gating mechanism into relation network to exploit global context information for adjusting per-channel modulation weights of local relation features. Extensive experiments on benchmark texture datasets and real scenarios demonstrate the above-par segmentation performance and robust generalization across domains of our proposed method.

Keywords

Cite

@article{arxiv.1905.06656,
  title  = {One-Shot Texture Retrieval with Global Context Metric},
  author = {Kai Zhu and Wei Zhai and Zheng-Jun Zha and Yang Cao},
  journal= {arXiv preprint arXiv:1905.06656},
  year   = {2020}
}

Comments

ijcai2019-lastest

R2 v1 2026-06-23T09:08:31.414Z