Optical Flow Estimation On Sintel Final
평가 지표
Average End-Point Error
평가 결과
이 벤치마크에서 각 모델의 성능 결과
비교 표
| 모델 이름 | Average End-Point Error |
|---|---|
| optical-flow-estimation-using-a-spatial | 8.36 |
| deep-equilibrium-optical-flow-estimation | 2.886 |
| selflow-self-supervised-learning-of-optical | 4.26 |
| fastflownet-a-lightweight-network-for-fast | 6.08 |
| fdflownet-fast-optical-flow-estimation-using | 5.11 |
| recurrent-partial-kernel-network-for | 2.657 |
| optical-flow-in-mostly-rigid-scenes | 5.38 |
| maskflownet-asymmetric-feature-matching-with | 4.17 |
| unifying-flow-stereo-and-depth-estimation | 2.37 |
| improved-cross-view-completion-pre-training | 2.436 |
| volumetric-correspondence-networks-for | 4.40 |
| iterative-residual-refinement-for-joint | 4.579 |
| perceiver-io-a-general-architecture-for | 2.42 |
| rethinking-raft-for-efficient-optical-flow | 2.60 |
| unsupervised-optical-flow-using-cost-function | 5.8 |
| global-matching-with-overlapping-attention | 2.648 |
| scopeflow-dynamic-scene-scoping-for-optical | 4.098 |
| learning-to-estimate-hidden-motions-with | 2.470 |
| continual-occlusions-and-optical-flow | 4.52 |
| a-lightweight-optical-flow-cnn-revisiting | 4.69 |
| liteflownet-a-lightweight-convolutional | 5.38 |
| raft-recurrent-all-pairs-field-transforms-for | 2.855 |
| dpflow-adaptive-optical-flow-estimation-with-1 | 1.975 |
| liteflownet3-resolving-correspondence | 4.45 |
| maskflownet-asymmetric-feature-matching-with | 4.38 |
| rapidflow-recurrent-adaptable-pyramids-with | 3.56 |
| liteflownet3-resolving-correspondence | 4.53 |