Method Details


Details for method 'ESPNetv2 (54 million FLOPs)'

 

Method overview

name ESPNetv2 (54 million FLOPs)
challenge pixel-level semantic labeling
details 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. The performance of our network is evaluated on three different tasks: (1) object classification, (2) semantic segmentation, and (3) language modeling. Experiments on these tasks, including image classification on the ImageNet and language modeling on the PenTree bank dataset, demonstrate the superior performance of our method over the state-of-the-art methods. Our network has better generalization properties than ShuffleNetv2 when tested on the MSCOCO multi-object classification task and the Cityscapes urban scene semantic segmentation task. Our experiments show that ESPNetv2 is much more power efficient than existing state-of-the-art efficient methods including ShuffleNets and MobileNets.
publication ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network
Sachin Mehta, Mohammad Rastegari, Linda Shapiro, Hannaneh Hajishirzi
https://arxiv.org/abs/1811.11431
project page / code https://github.com/sacmehta/ESPNetv2
used Cityscapes data fine annotations
used external data ImageNet
runtime 0.007 s
NVIDIA TitanX
subsampling 2
submission date November, 2018
previous submissions

 

Average results

Metric Value
IoU Classes 54.6644
iIoU Classes 27.9719
IoU Categories 78.7408
iIoU Categories 59.4566

 

Class results

Class IoU iIoU
road 94.4974 -
sidewalk 67.8543 -
building 84.8879 -
wall 33.3857 -
fence 32.857 -
pole 35.378 -
traffic light 38.6435 -
traffic sign 48.2312 -
vegetation 88.4361 -
terrain 61.9146 -
sky 91.0038 -
person 63.2616 41.0694
rider 40.8039 16.3729
car 87.2115 78.6925
truck 32.5849 11.9599
bus 38.1062 21.1901
train 15.0069 10.4942
motorcycle 35.957 11.1943
bicycle 48.6013 32.8021

 

Category results

Category IoU iIoU
flat 94.7853 -
nature 87.8959 -
object 43.1427 -
sky 91.0038 -
construction 84.8535 -
human 64.5251 42.4574
vehicle 84.979 76.4559

 

Links

Download results as .csv file

Benchmark page