MaxpoolNMS: Getting Rid of NMS Bottlenecks in Two-Stage Object Detectors

Lile Cai, Bin Zhao, Zhe Wang, Jie Lin, Chuan Sheng Foo, Mohamed Sabry Aly, Vijay Chandrasekhar; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 9356-9364

Abstract


Modern convolutional object detectors have improved the detection accuracy significantly, which in turn inspired the development of dedicated hardware accelerators to achieve real-time performance by exploiting inherent parallelism in the algorithm. Non-maximum suppression (NMS) is an indispensable operation in object detection. In stark contrast to most operations, the commonly-adopted GreedyNMS algorithm does not foster parallelism, which can be a major performance bottleneck. In this paper, we introduce MaxpoolNMS, a parallelizable alternative to the NMS algorithm, which is based on max-pooling classification score maps. By employing a novel multi-scale multi-channel max-pooling strategy, our method is 20x faster than GreedyNMS while simultaneously achieves comparable accuracy, when quantified across various benchmarking datasets, i.e., MS COCO, KITTI and PASCAL VOC. Furthermore, our method is better suited for hardware-based acceleration than GreedyNMS.

Related Material


[pdf]
[bibtex]
@InProceedings{Cai_2019_CVPR,
author = {Cai, Lile and Zhao, Bin and Wang, Zhe and Lin, Jie and Foo, Chuan Sheng and Aly, Mohamed Sabry and Chandrasekhar, Vijay},
title = {MaxpoolNMS: Getting Rid of NMS Bottlenecks in Two-Stage Object Detectors},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}
}