ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network

CVPR 2019 Sachin MehtaMohammad RastegariLinda ShapiroHannaneh Hajishirzi

We introduce a light-weight, power efficient, and general purpose convolutional neural network, ESPNetv2, for modeling visual and sequential data. Our network uses group point-wise and depth-wise dilated separable convolutions to learn representations from a large effective receptive field with fewer FLOPs and parameters... (read more)

PDF Abstract

Evaluation results from the paper

Task Dataset Model Metric name Metric value Global rank Compare
Semantic Segmentation Cityscapes test ESPNetv2 Mean IoU (class) 66.2% # 37
Object Detection COCO test-dev ESPNetv2-512 box AP 26.0 # 59
Image Classification ImageNet ESPNetv2 Top 1 Accuracy 74.9% # 77
Object Detection PASCAL VOC 2007 ESPNetv2-512 MAP 68.2% # 19
Semantic Segmentation PASCAL VOC 2012 test ESPNetv2 Mean IoU 68.0% # 25