Method Details

Details for method 'LevelSet R-CNN [fine-only]'


Method overview

name LevelSet R-CNN [fine-only]
challenge instance-level semantic labeling
details Obtaining precise instance segmentation masks is of high importance in many modern applications such as robotic manipulation and autonomous driving. Currently, many state of the art models are based on the Mask R-CNN framework which, while very powerful, outputs masks at low resolutions which could result in imprecise boundaries. On the other hand, classic variational methods for segmentation impose desirable global and local data and geometry constraints on the masks by optimizing an energy functional. While mathematically elegant, their direct dependence on good initialization, non-robust image cues and manual setting of hyperparameters renders them unsuitable for modern applications. We propose LevelSet R-CNN, which combines the best of both worlds by obtaining powerful feature representations that are combined in an end-to-end manner with a variational segmentation framework. We demonstrate the effectiveness of our approach on COCO and Cityscapes datasets.
publication LevelSet R-CNN: A Deep Variational Method for Instance Segmentation
Namdar Homayounfar*, Yuwen Xiong*, Justin Liang*, Wei-Chiu Ma, Raquel Urtasun
ECCV 2020
project page / code
used Cityscapes data fine annotations
used external data ImageNet
runtime n/a
subsampling no
submission date March, 2020
previous submissions


Average results

Metric Value
AP 33.3395
AP50% 58.2467
AP100m 47.4733
AP50m 50.3464


Class results

Class AP AP50% AP100m AP50m
person 36.9686 68.3193 55.1506 55.3092
rider 29.2809 66.605 42.5295 42.8485
car 54.6191 77.2733 74.8934 77.9999
truck 30.4487 42.6904 41.6978 48.3821
bus 39.3659 55.5786 57.5106 62.1405
train 30.245 53.4421 44.4707 51.3404
motorcycle 25.4801 53.2664 33.4235 34.702
bicycle 20.3077 48.7985 30.1099 30.049



Download results as .csv file

Benchmark page