Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation

CVPR 2020 Liang LiuJiangning ZhangRuifei HeYong LiuYabiao WangYing TaiDonghao LuoChengjie WangJilin LiFeiyue Huang

Unsupervised learning of optical flow, which leverages the supervision from view synthesis, has emerged as a promising alternative to supervised methods. However, the objective of unsupervised learning is likely to be unreliable in challenging scenes... (read more)

PDF Abstract CVPR 2020 PDF CVPR 2020 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Optical Flow Estimation KITTI 2012 unsupervised ARFlow-MV Average End-Point Error 1.5 # 1
Optical Flow Estimation KITTI 2015 unsupervised ARFlow-MV Fl-all 11.79 # 1
Optical Flow Estimation Sintel Clean unsupervised ARFlow-MV Average End-Point Error 4.49 # 1
Optical Flow Estimation Sintel Final unsupervised ARFlow-MV Average End-Point Error 5.67 # 1

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet