LiteFlowNet3: Resolving Correspondence Ambiguity for More Accurate Optical Flow Estimation

Open paper
Benchmark Model Rank Results
optical-flow-estimation-on-kitti-2012LiteFlowNet3#4Average End-Point Error: 1.3Noc: 0.7
optical-flow-estimation-on-kitti-2012LiteFlowNet3-S#5Average End-Point Error: 1.3Noc: 0.7
optical-flow-estimation-on-kitti-2015LiteFlowNet3-S#28Fl-all: 7.22Fl-fg: 6.96
optical-flow-estimation-on-kitti-2015LiteFlowNet3#29Fl-all: 7.34Fl-fg: 7.75
optical-flow-estimation-on-sintel-cleanLiteFlowNet3#28Average End-Point Error: 2.99
optical-flow-estimation-on-sintel-cleanLiteFlowNet3-S#29Average End-Point Error: 3.03
optical-flow-estimation-on-sintel-finalLiteFlowNet3#29Average End-Point Error: 4.45
optical-flow-estimation-on-sintel-finalLiteFlowNet3-S#30Average End-Point Error: 4.53

Rank counts only results with a code link.