Optical Flow Estimation On Sintel Clean

Metrics

Average End-Point Error

Results

Performance results of various models on this benchmark

Model Name	Average End-Point Error	Paper Title	Repository
FDFlowNet-ft	3.71	FDFlowNet: Fast Optical Flow Estimation using a Deep Lightweight Network	-
RAPIDFlow	2.03	RAPIDFlow: Recurrent Adaptable Pyramids with Iterative Decoding for Efficient Optical Flow Estimation
ScopeFlow	3.592	ScopeFlow: Dynamic Scene Scoping for Optical Flow	-
LiteFlowNet2-ft	3.48	A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularization	-
MaskFlownet	2.52	MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask	-
DEQ-Flow-H	1.519	Deep Equilibrium Optical Flow Estimation	-
GMFlow	1.03	Unifying Flow, Stereo and Depth Estimation	-
RPKNet	1.315	Recurrent Partial Kernel Network for Efficient Optical Flow Estimation
RAFT (warm-start)	1.609	RAFT: Recurrent All-Pairs Field Transforms for Optical Flow	-
IRR-PWC	3.84	Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation	-
SelFlow	3.74	SelFlow: Self-Supervised Learning of Optical Flow	-
GMA	1.388	Learning to Estimate Hidden Motions with Global Motion Aggregation	-
LiteFlowNet3-S	3.03	LiteFlowNet3: Resolving Correspondence Ambiguity for More Accurate Optical Flow Estimation	-
Ef-RAFT	1.27	Rethinking RAFT for Efficient Optical Flow	-
ProFlow+ft	2.82	ProFlow: Learning to Predict Optical Flow	-
UnrolledCost	4.69	Cost Function Unrolling in Unsupervised Optical Flow	-
FlowNet2	3.96	FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks	-
Perceiver IO	1.81	Perceiver IO: A General Architecture for Structured Inputs & Outputs	-
FlowFormer	1.16	FlowFormer: A Transformer Architecture for Optical Flow	-
FastFlowNet-ft	4.89	FastFlowNet: A Lightweight Network for Fast Optical Flow Estimation	-

0 of 28 row(s) selected.