MegDet: A Large Mini-Batch Object Detector
Chao Peng, Tete Xiao, Zeming Li et al. · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · 2018
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. However, mini-batch size, a key factor for the training of deep neural networks, has not been well studied for object detection. 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 submission (mmAP 52.5%) to COCO 2017 Challenge, where we won the 1st place of Detection task.
Cited by 302 works. doi:10.1109/cvpr.2018.00647 Open-access PDF
Authors
References
- Deep Residual Learning for Image Recognition
- ImageNet: A large-scale hierarchical image database
- Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- Microsoft COCO: Common Objects in Context
- You Only Look Once: Unified, Real-Time Object Detection
- Fast R-CNN
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- SSD: Single Shot MultiBox Detector
- Selective Search for Object Recognition