Compact Global Descriptor for Neural Networks

23 Jul 2019Xiangyu HeKe ChengQiang ChenQinghao HuPeisong WangJian Cheng

Long-range dependencies modeling, widely used in capturing spatiotemporal correlation, has shown to be effective in CNN dominated computer vision tasks. Yet neither stacks of convolutional operations to enlarge receptive fields nor recent nonlocal modules is computationally efficient... (read more)

PDF Abstract

Evaluation results from the paper

Task Dataset Model Metric name Metric value Global rank Compare
Object Detection COCO test-dev Faster R-CNN + FPN + CGD box AP 37.9 # 47
Object Detection COCO test-dev MobielNet-v1-SSD-300x300+CGD box AP 21.4 # 61
Image Classification ImageNet MobileNet-224 (CGD) Top 1 Accuracy 72.56% # 85
Image Classification ImageNet MobileNet-224 (CGD) Top 5 Accuracy 90.92% # 72
Object Detection PASCAL VOC 2007 MobielNet-v1-SSD-300x300+CGD MAP 74.8 # 13