Method

MaskFlownet-S [MaskFlownet-S]
https://github.com/microsoft/MaskFlownet

Submitted on 15 Nov. 2019 07:24 by
Shengyu Zhao (Tsinghua University)

Running time:0.03 s
Environment:NVIDIA TITAN Xp

Method Description:
Feature warping is a core technique in optical
flow estimation; however, the ambiguity caused by
occluded areas during warping is a major problem
that remains unsolved. In this paper, we propose
an asymmetric occlusion-aware feature matching
module, which can learn a rough occlusion mask
that filters useless (occluded) areas immediately
after feature warping without any explicit
supervision. The proposed module can be easily
integrated into end-to-end network architectures
and enjoys performance gains while introducing
negligible computational cost. The learned
occlusion mask can be further fed into a
subsequent network cascade with dual feature
pyramids with which we achieve state-of-the-art
performance. At the time of submission, our
method, called MaskFlownet, surpasses all
published optical flow methods on the MPI Sintel,
KITTI 2012 and 2015 benchmarks.
Parameters:
CTSK
Latex Bibtex:
@inproceedings{zhao2020maskflownet,
author = {Zhao, Shengyu and Sheng, Yilun and Dong,
Yue and Chang, Eric I-Chao and Xu, Yan},
title = {MaskFlownet: Asymmetric Feature Matching
with Learnable Occlusion Mask},
booktitle = {Proceedings of the IEEE Conference on
Computer Vision and Pattern Recognition (CVPR)},
year = {2020}
}

Detailed Results

This page provides detailed results for the method(s) selected. For the first 20 test images, the percentage of erroneous pixels is depicted in the table. We use the error metric described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), which considers a pixel to be correctly estimated if the disparity or flow end-point error is <3px or <5% (for scene flow this criterion needs to be fulfilled for both disparity maps and the flow map). Underneath, the left input image, the estimated results and the error maps are shown (for disp_0/disp_1/flow/scene_flow, respectively). The error map uses the log-color scale described in Object Scene Flow for Autonomous Vehicles (CVPR 2015), depicting correct estimates (<3px or <5% error) in blue and wrong estimates in red color tones. Dark regions in the error images denote the occluded pixels which fall outside the image boundaries. The false color maps of the results are scaled to the largest ground truth disparity values / flow magnitudes.

Test Set Average

Error Fl-bg Fl-fg Fl-all
All / All 6.53 8.21 6.81
All / Est 6.53 8.21 6.81
Noc / All 4.03 5.39 4.27
Noc / Est 4.03 5.39 4.27
This table as LaTeX

Test Image 0

Error Fl-bg Fl-fg Fl-all
All / All 2.90 17.47 4.90
All / Est 2.90 17.47 4.90
Noc / All 2.95 17.47 5.15
Noc / Est 2.95 17.47 5.15
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 1

Error Fl-bg Fl-fg Fl-all
All / All 2.29 24.77 4.80
All / Est 2.29 24.77 4.80
Noc / All 2.23 24.77 5.01
Noc / Est 2.23 24.77 5.01
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 2

Error Fl-bg Fl-fg Fl-all
All / All 8.13 3.15 7.89
All / Est 8.13 3.15 7.89
Noc / All 6.86 3.15 6.65
Noc / Est 6.86 3.15 6.65
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 3

Error Fl-bg Fl-fg Fl-all
All / All 11.18 9.04 10.98
All / Est 11.18 9.04 10.98
Noc / All 9.15 7.52 9.01
Noc / Est 9.15 7.52 9.01
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 4

Error Fl-bg Fl-fg Fl-all
All / All 4.86 13.62 6.31
All / Est 4.86 13.62 6.31
Noc / All 4.34 12.23 5.81
Noc / Est 4.34 12.23 5.81
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 5

Error Fl-bg Fl-fg Fl-all
All / All 3.65 0.20 3.34
All / Est 3.65 0.20 3.34
Noc / All 2.91 0.20 2.63
Noc / Est 2.91 0.20 2.63
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 6

Error Fl-bg Fl-fg Fl-all
All / All 4.05 0.48 3.67
All / Est 4.05 0.48 3.67
Noc / All 2.17 0.48 1.97
Noc / Est 2.17 0.48 1.97
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 7

Error Fl-bg Fl-fg Fl-all
All / All 1.94 17.39 4.96
All / Est 1.94 17.39 4.96
Noc / All 1.94 15.29 4.43
Noc / Est 1.94 15.29 4.43
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 8

Error Fl-bg Fl-fg Fl-all
All / All 0.88 3.12 1.29
All / Est 0.88 3.12 1.29
Noc / All 0.88 3.12 1.29
Noc / Est 0.88 3.12 1.29
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 9

Error Fl-bg Fl-fg Fl-all
All / All 0.93 18.87 5.51
All / Est 0.93 18.87 5.51
Noc / All 0.93 18.87 5.51
Noc / Est 0.93 18.87 5.51
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 10

Error Fl-bg Fl-fg Fl-all
All / All 1.90 4.05 2.39
All / Est 1.90 4.05 2.39
Noc / All 1.85 4.05 2.42
Noc / Est 1.85 4.05 2.42
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 11

Error Fl-bg Fl-fg Fl-all
All / All 2.40 2.82 2.48
All / Est 2.40 2.82 2.48
Noc / All 2.49 2.81 2.55
Noc / Est 2.49 2.81 2.55
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 12

Error Fl-bg Fl-fg Fl-all
All / All 3.04 2.42 3.00
All / Est 3.04 2.42 3.00
Noc / All 1.46 2.42 1.54
Noc / Est 1.46 2.42 1.54
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 13

Error Fl-bg Fl-fg Fl-all
All / All 3.55 2.33 3.40
All / Est 3.55 2.33 3.40
Noc / All 1.56 1.91 1.61
Noc / Est 1.56 1.91 1.61
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 14

Error Fl-bg Fl-fg Fl-all
All / All 3.54 2.53 3.52
All / Est 3.54 2.53 3.52
Noc / All 2.52 2.53 2.52
Noc / Est 2.52 2.53 2.52
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 15

Error Fl-bg Fl-fg Fl-all
All / All 10.30 4.60 9.79
All / Est 10.30 4.60 9.79
Noc / All 7.51 4.60 7.17
Noc / Est 7.51 4.60 7.17
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 16

Error Fl-bg Fl-fg Fl-all
All / All 15.05 4.96 13.57
All / Est 15.05 4.96 13.57
Noc / All 8.04 4.96 7.48
Noc / Est 8.04 4.96 7.48
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 17

Error Fl-bg Fl-fg Fl-all
All / All 5.25 3.29 5.05
All / Est 5.25 3.29 5.05
Noc / All 3.85 3.29 3.78
Noc / Est 3.85 3.29 3.78
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 18

Error Fl-bg Fl-fg Fl-all
All / All 17.20 97.85 55.52
All / Est 17.20 97.85 55.52
Noc / All 11.44 93.15 33.97
Noc / Est 11.44 93.15 33.97
This table as LaTeX

Input Image

Flow Result

Flow Error


Test Image 19

Error Fl-bg Fl-fg Fl-all
All / All 3.86 3.92 3.87
All / Est 3.86 3.92 3.87
Noc / All 3.49 3.92 3.55
Noc / Est 3.49 3.92 3.55
This table as LaTeX

Input Image

Flow Result

Flow Error




eXTReMe Tracker