• [Soft-NMS] N. Bodla, B. Singh, R. Chellappa, L. S. Davis, "Soft-NMS Improving Object Detection With One Line of Code", ICCV, 2017.

• [Non-maximum Weighted - NMW] H. Zhou, Z. Li, C. Ning, J. Tang, "CAD: Scale-Invariant Framework for Real-Time Object Detection", Proceedings of the IEEE International Conference on Computer Vision, ICCV, 2017

• [WBF] R. Solovyev, W. Wang, T. Gabruseva, "Weighted boxes fusion: Ensembling boxes from different object detection models", Image and Vision Computing, 2021.

• Hosang, J.H., Benenson, R., Schiele, B.: Learning non-maximum suppression. In: CVPR (2017).

Unfilter

• Non Maximum Suppresion (NMS): 

  • Used to filter those BBoxes that badly predict the same object.

  • Only retain the candidate BBoxes with higher response. 

• Greedy NMS [19]:

  • Added classification confidence score in R-CNN as a reference.

• Soft NMS [1]: 

  • Add IoU score to consider the problem of the occlusion of an object (cause the degradation of confidence score). 

• DIoU NMS [99]: 

  • Add the center point distance to the BBox screening process.