Learning to Filter Object Detections

@inproceedings{Prokudin2017LearningTF,
  title={Learning to Filter Object Detections},
  author={S. Prokudin and D. Kappler and S. Nowozin and P. Gehler},
  booktitle={GCPR},
  year={2017}
}
Most object detection systems consist of three stages. First, a set of individual hypotheses for object locations is generated using a proposal generating algorithm. Second, a classifier scores every generated hypothesis independently to obtain a multi-class prediction. Finally, all scored hypotheses are filtered via a non-differentiable and decoupled non-maximum suppression (NMS) post-processing step. In this paper, we propose a filtering network (FNet), a method which replaces NMS with a… Expand
2 Citations
GrooMeD-NMS: Grouped Mathematically Differentiable NMS for Monocular 3D Object Detection
  • Abhinav Kumar, G. Brazil, Xiaoming Liu
  • Computer Science
  • ArXiv
  • 2021
  • PDF
Daedalus: Breaking Non-Maximum Suppression in Object Detection via Adversarial Examples
  • 20
  • PDF

References

SHOWING 1-10 OF 18 REFERENCES
You Only Look Once: Unified, Real-Time Object Detection
  • 11,381
  • PDF
End-to-end integration of a Convolutional Network, Deformable Parts Model and non-maximum suppression
  • Li Wan, D. Eigen, R. Fergus
  • Computer Science
  • 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2015
  • 80
  • PDF
Non-maximum Suppression for Object Detection by Passing Messages Between Windows
  • 106
  • PDF
End-to-End Training of Object Class Detectors for Mean Average Precision
  • 57
  • Highly Influential
  • PDF
A Convnet for Non-maximum Suppression
  • 43
  • PDF
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
  • 23,303
  • Highly Influential
  • PDF
End-to-End People Detection in Crowded Scenes
  • 280
  • PDF
A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection
  • 903
  • Highly Influential
  • PDF
Pedestrian Detection: An Evaluation of the State of the Art
  • 2,494
  • PDF
The Pascal Visual Object Classes (VOC) Challenge
  • 9,346
  • Highly Influential
  • PDF
...
1
2
...