Home News Latest Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Optical Flow Estimation On Kitti 2015 Train

Metrics

EPE

F1-all

Results

Performance results of various models on this benchmark

Model Name	EPE	F1-all	Paper Title	Repository
VCN	8.36	25.1	Volumetric Correspondence Networks for Optical Flow
CRAFT	4.88	17.5	CRAFT: Cross-Attentional Flow Transformer for Robust Optical Flow
RAFT	5.04	17.4	RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
PWC-Net	10.35	33.7	PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume
FastFlowNet	12.24	33.1	FastFlowNet: A Lightweight Network for Fast Optical Flow Estimation
GMA	4.69	17.1	Learning to Estimate Hidden Motions with Global Motion Aggregation
FlowNet2	10.08	30.0	FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks
FlowFormer	4.09	14.7	FlowFormer: A Transformer Architecture for Optical Flow
HD3	13.17	24.0	Hierarchical Discrete Distribution Decomposition for Match Density Estimation
Ef-RAFT	4.83	16.45	Rethinking RAFT for Efficient Optical Flow
RPKNet	3.79	13.0	Recurrent Partial Kernel Network for Efficient Optical Flow Estimation
RAPIDFlow	5.87	17.7	RAPIDFlow: Recurrent Adaptable Pyramids with Iterative Decoding for Efficient Optical Flow Estimation
DEQ-Flow	3.76	13.0	Deep Equilibrium Optical Flow Estimation
GMFlowNet	4.24	15.4	Global Matching with Overlapping Attention for Optical Flow Estimation
MaskFlowNet	-	23.1	MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask
SCV	6.80	19.3	Learning Optical Flow from a Few Matches
SeparableFlow	4.60	15.9	Separable Flow: Learning Motion Cost Volumes for Optical Flow Estimation
DPFlow	3.37	11.1	-	-

0 of 18 row(s) selected.