SNIPER: Efficient Multi-Scale Training

NeurIPS 2018 Bharat Singh • Mahyar Najibi • Larry S. Davis

We present SNIPER, an algorithm for performing efficient multi-scale training in instance level visual recognition tasks. As SNIPER operates on resampled low resolution chips (512x512 pixels), it can have a batch size as large as 20 on a single GPU even with a ResNet-101 backbone. Our implementation based on Faster-RCNN with a ResNet-101 backbone obtains an mAP of 47.6% on the COCO dataset for bounding box detection and can process 5 images per second during inference with a single GPU.

PDF Abstract


Task Dataset Model Metric name Metric value Global rank Compare
Object Detection COCO SNIPER Bounding Box AP 47.6 # 1
Object Detection PASCAL VOC 2007 SNIPER MAP 86.9% # 1