MegDet: A Large Mini-Batch Object Detector

Chao Peng, Tete Xiao, Zeming Li, Yuning Jiang, Xiangyu Zhang, Kai Jia, Gang Yu, Jian Sun; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 6181-6189


The development of object detection in the era of deep learning, from R-CNN [11], Fast/Faster R-CNN [10, 31] to recent Mask R-CNN [14] and RetinaNet [24], mainly come from novel network, new framework, or loss design. How- ever, mini-batch size, a key factor for the training of deep neural networks, has not been well studied for object detec- tion. In this paper, we propose a Large Mini-Batch Object Detector (MegDet) to enable the training with a large mini- batch size up to 256, so that we can effectively utilize at most 128 GPUs to significantly shorten the training time. Technically, we suggest a warmup learning rate policy and Cross-GPU Batch Normalization, which together allow us to successfully train a large mini-batch detector in much less time (e.g., from 33 hours to 4 hours), and achieve even better accuracy. The MegDet is the backbone of our sub- mission (mmAP 52.5%) to COCO 2017 Challenge, where we won the 1st place of Detection task.

Related Material

[pdf] [arXiv] [video]
author = {Peng, Chao and Xiao, Tete and Li, Zeming and Jiang, Yuning and Zhang, Xiangyu and Jia, Kai and Yu, Gang and Sun, Jian},
title = {MegDet: A Large Mini-Batch Object Detector},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2018}