FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

Benchmark Model Rank Results
dense-pixel-correspondence-estimation-onFlowNet2#5Viewpoint I AEPE: 5.99Viewpoint II AEPE: 15.55
optical-flow-estimation-on-kitti-2015-trainFlowNet2#17F1-all: 30.0EPE: 10.08
optical-flow-estimation-on-sintel-cleanFlowNet2#22Average End-Point Error: 3.96
optical-flow-estimation-on-springFlowNet2#61px total: 6.710