Scipediacontent: Scipediacontent moved page Draft Content 342042623 to Chellappa et al 2017a

2021-01-28T16:16:13Z

Scipediacontent moved page Draft Content 342042623 to Chellappa et al 2017a

Scipediacontent: Created page with " == Abstract == Non-maximum suppression is an integral part of the object detection pipeline. First, it sorts all detection boxes on the basis of their scores. The detection..."

2021-01-28T16:16:08Z

Created page with " == Abstract == Non-maximum suppression is an integral part of the object detection pipeline. First, it sorts all detection boxes on the basis of their scores. The detection..."

New page

== Abstract ==

Non-maximum suppression is an integral part of the object detection pipeline. First, it sorts all detection boxes on the basis of their scores. The detection box M with the maximum score is selected and all other detection boxes with a significant overlap (using a pre-defined threshold) with M are suppressed. This process is recursively applied on the remaining boxes. As per the design of the algorithm, if an object lies within the predefined overlap threshold, it leads to a miss. To this end, we propose Soft-NMS, an algorithm which decays the detection scores of all other objects as a continuous function of their overlap with M. Hence, no object is eliminated in this process. Soft-NMS obtains consistent improvements for the coco-style mAP metric on standard datasets like PASCAL VOC 2007 (1.7% for both R-FCN and Faster-RCNN) and MS-COCO (1.3% for R-FCN and 1.1% for Faster-RCNN) by just changing the NMS algorithm without any additional hyper-parameters. Using Deformable-RFCN, Soft-NMS improves state-of-the-art in object detection from 39.8% to 40.9% with a single model. Further, the computational complexity of Soft-NMS is the same as traditional NMS and hence it can be efficiently implemented. Since Soft-NMS does not require any extra training and is simple to implement, it can be easily integrated into any object detection pipeline. Code for Soft-NMS is publicly available on GitHub (http://bit.ly/2nJLNMu).

Comment: ICCV 2017 camera ready version

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1704.04503 http://arxiv.org/abs/1704.04503]

* [http://arxiv.org/pdf/1704.04503 http://arxiv.org/pdf/1704.04503]

* [http://dx.doi.org/10.1109/iccv.2017.593 http://dx.doi.org/10.1109/iccv.2017.593]

* [http://xplorestaging.ieee.org/ielx7/8234942/8237262/08237855.pdf?arnumber=8237855 http://xplorestaging.ieee.org/ielx7/8234942/8237262/08237855.pdf?arnumber=8237855],
: [http://dx.doi.org/10.1109/iccv.2017.593 http://dx.doi.org/10.1109/iccv.2017.593]

* [https://dblp.uni-trier.de/db/conf/iccv/iccv2017.html#BodlaSCD17 https://dblp.uni-trier.de/db/conf/iccv/iccv2017.html#BodlaSCD17],
: [https://ieeexplore.ieee.org/document/8237855 https://ieeexplore.ieee.org/document/8237855],
: [https://arxiv.org/pdf/1704.04503 https://arxiv.org/pdf/1704.04503],
: [https://www.arxiv-vanity.com/papers/1704.04503 https://www.arxiv-vanity.com/papers/1704.04503],
: [http://ieeexplore.ieee.org/document/8237855 http://ieeexplore.ieee.org/document/8237855],
: [https://doi.org/10.1109/ICCV.2017.593 https://doi.org/10.1109/ICCV.2017.593],
: [https://academic.microsoft.com/#/detail/2964121718 https://academic.microsoft.com/#/detail/2964121718]

← Older revision	Revision as of 16:16, 28 January 2021
(No difference)

Chellappa et al 2017a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 342042623 to Chellappa et al 2017a

Scipediacontent: Created page with " == Abstract == Non-maximum suppression is an integral part of the object detection pipeline. First, it sorts all detection boxes on the basis of their scores. The detection..."