Corpus ID: 219573791

Quasi-Dense Instance Similarity Learning

@article{Pang2020QuasiDenseIS,
  title={Quasi-Dense Instance Similarity Learning},
  author={Jiangmiao Pang and L. Qiu and H. Chen and Q. Li and Trevor Darrell and F. Yu},
  journal={ArXiv},
  year={2020},
  volume={abs/2006.06664}
}
Similarity metrics for instances have drawn much attention, due to their importance for computer vision problems such as object tracking. However, existing methods regard object similarity learning as a post-hoc stage after object detection and only use sparse ground truth matching as the training objective. This process ignores the majority of the regions on the images. In this paper, we present a simple yet effective quasi-dense matching method to learn instance similarity from hundreds of… Expand
5 Citations
Learning to Track Instances without Video Annotations
  • PDF
Crossover Learning for Fast Online Video Instance Segmentation
  • Shusheng Yang, Yuxin Fang, +5 authors Wenyu Liu
  • Computer Science
  • ArXiv
  • 2021
  • PDF
Local Metrics for Multi-Object Tracking
  • Highly Influenced
  • PDF

References

SHOWING 1-10 OF 63 REFERENCES
RepMet: Representative-Based Metric Learning for Classification and Few-Shot Object Detection
  • 91
  • PDF
Few-Shot Object Detection via Feature Reweighting
  • 112
  • PDF
Unsupervised Feature Learning via Non-parametric Instance Discrimination
  • 484
  • PDF
Learning by Tracking: Siamese CNN for Robust Target Association
  • 247
  • PDF
Meta R-CNN: Towards General Solver for Instance-Level Low-Shot Learning
  • 54
  • Highly Influential
  • PDF
Microsoft COCO: Common Objects in Context
  • 13,183
  • PDF
Mask R-CNN
  • 5,787
  • Highly Influential
Meta-Learning to Detect Rare Objects
  • Yu-Xiong Wang, D. Ramanan, M. Hebert
  • Computer Science
  • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
  • 2019
  • 45
  • PDF
Feature Pyramid Networks for Object Detection
  • 5,928
  • PDF
...
1
2
3
4
5
...