awesome-optical-flow
This is a list of awesome paper about optical flow and related work.
https://github.com/hzwer/awesome-optical-flow
Last synced: 5 days ago
JSON representation
-
Before 2020
-
Classical Estimation Methods
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- DeepFlow: Large Displacement Optical Flow with Deep Matching
- Optical Flow Estimation with Channel Constancy
- An iterative image registration technique with an application to stereo vision
- Determining optical flow
- Motion Detail Preserving Optical Flow Estimation
- Secrets of Optical Flow Estimation and Their Principles
- S2F: Slow-To-Fast Interpolator Flow
- An iterative image registration technique with an application to stereo vision
- Determining optical flow
- Motion Detail Preserving Optical Flow Estimation
- Secrets of Optical Flow Estimation and Their Principles
- DeepFlow: Large Displacement Optical Flow with Deep Matching
- Optical Flow Estimation with Channel Constancy
- S2F: Slow-To-Fast Interpolator Flow
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
- Optical Flow Estimation with Channel Constancy
-
Others
- On the Integration of Optical Flow and Action Recognition
- Spatially Smooth Optical Flow for Video Stabilization
- Volumetric Correspondence Networks for Optical Flow - y/VCN) 
- Volumetric Correspondence Networks for Optical Flow - y/VCN) 
- Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation
- Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation
- PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume - Net](https://github.com/NVlabs/PWC-Net)  | [pytorch-pwc](https://github.com/sniklaus/pytorch-pwc) 
- LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow Estimation - liteflownet](https://github.com/sniklaus/pytorch-liteflownet) 
- FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks - pytorch](https://github.com/NVIDIA/flownet2-pytorch)  <br> [flownet2](https://github.com/lmb-freiburg/flownet2)  <br> [flownet2-tf](https://github.com/sampepose/flownet2-tf) 
- Optical Flow Estimation using a Spatial Pyramid Network - spynet](https://github.com/sniklaus/pytorch-spynet) 
- PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume - Net](https://github.com/NVlabs/PWC-Net)  | [pytorch-pwc](https://github.com/sniklaus/pytorch-pwc) 
- LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow Estimation - liteflownet](https://github.com/sniklaus/pytorch-liteflownet) 
- FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks - pytorch](https://github.com/NVIDIA/flownet2-pytorch)  <br> [flownet2](https://github.com/lmb-freiburg/flownet2)  <br> [flownet2-tf](https://github.com/sampepose/flownet2-tf) 
- Optical Flow Estimation using a Spatial Pyramid Network - spynet](https://github.com/sniklaus/pytorch-spynet) 
- FlowNet: Learning Optical Flow with Convolutional Networks
- DDFlow: Learning Optical Flow with Unlabeled Data Distillation
- FlowNet: Learning Optical Flow with Convolutional Networks
- DDFlow: Learning Optical Flow with Unlabeled Data Distillation
- SelFlow: Self-Supervised Learning of Optical Flow
- Unsupervised Deep Epipolar Flow for Stationary or Dynamic Scenes
- Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose
- RainFlow: Optical Flow under Rain Streaks and Rain Veiling Effect
- Robust Optical Flow Estimation in Rainy Scenes
- SelFlow: Self-Supervised Learning of Optical Flow
- Unsupervised Deep Epipolar Flow for Stationary or Dynamic Scenes
- Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose
- RainFlow: Optical Flow under Rain Streaks and Rain Veiling Effect
- Robust Optical Flow Estimation in Rainy Scenes
- Quadratic Video Interpolation
- Depth-Aware Video Frame Interpolation
- Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolation - SloMo](https://github.com/avinashpaliwal/Super-SloMo) 
- Video Frame Synthesis using Deep Voxel Flow - flow](https://github.com/liuziwei7/voxel-flow)  | [pytorch-voxel-flow](https://github.com/lxx1991/pytorch-voxel-flow) 
- DVC: An End-to-end Deep Video Compression Framework
- Quadratic Video Interpolation
- Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolation - SloMo](https://github.com/avinashpaliwal/Super-SloMo) 
- Video Frame Synthesis using Deep Voxel Flow - flow](https://github.com/liuziwei7/voxel-flow)  | [pytorch-voxel-flow](https://github.com/lxx1991/pytorch-voxel-flow) 
- DVC: An End-to-end Deep Video Compression Framework
- SegFlow: Joint Learning for Video Object Segmentation and Optical Flow
- End-to-end Flow Correlation Tracking with Spatial-temporal Attention
- Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition - Flow-Guided-Feature](https://github.com/kevin-ssy/Optical-Flow-Guided-Feature) 
- On the Integration of Optical Flow and Action Recognition
- Spatially Smooth Optical Flow for Video Stabilization
- SegFlow: Joint Learning for Video Object Segmentation and Optical Flow
- End-to-end Flow Correlation Tracking with Spatial-temporal Attention
- Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition - Flow-Guided-Feature](https://github.com/kevin-ssy/Optical-Flow-Guided-Feature) 
-
-
Optical Flow
-
Supervised Models
- MemFlow: Optical Flow Estimation and Prediction with Memory
- DistractFlow: Improving Optical Flow Estimation via Realistic Distractions and Pseudo-Labeling
- Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimation
- SKFlow: Learning Optical Flow with Super Kernels
- Disentangling architecture and training for optical flow - research/opticalflow-autoflow) |
- FlowFormer: A Transformer Architecture for Optical Flow - Official/) |
- Learning Optical Flow with Kernel Patch Attention - research/KPAFlow) |
- GMFlow: Learning Optical Flow via Global Matching
- SKFlow: Learning Optical Flow with Super Kernels
- GMFlow: Learning Optical Flow via Global Matching
- Deep Equilibrium Optical Flow Estimation - flow](https://github.com/locuslab/deq-flow) |
- High-Resolution Optical Flow from 1D Attention and Correlation
- Disentangling architecture and training for optical flow - research/opticalflow-autoflow) |
- FlowFormer: A Transformer Architecture for Optical Flow - Official/) |
- Learning Optical Flow with Kernel Patch Attention - research/KPAFlow) |
- Learning to Estimate Hidden Motions with Global Motion Aggregation
- Learning Optical Flow from a Few Matches
- Detail Preserving Coarse-to-Fine Matching for Stereo Matching and Optical Flow
- RAFT: Recurrent All Pairs Field Transforms for Optical Flow - vl/RAFT) 
- MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask
- ScopeFlow: Dynamic Scene Scoping for Optical Flow
- A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularization
- MemFlow: Optical Flow Estimation and Prediction with Memory
- DistractFlow: Improving Optical Flow Estimation via Realistic Distractions and Pseudo-Labeling
- Learning Optical Flow from a Few Matches
- Detail Preserving Coarse-to-Fine Matching for Stereo Matching and Optical Flow
- RAFT: Recurrent All Pairs Field Transforms for Optical Flow - vl/RAFT) 
- MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask
- ScopeFlow: Dynamic Scene Scoping for Optical Flow
- A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularization
- Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimation
- Deep Equilibrium Optical Flow Estimation - flow](https://github.com/locuslab/deq-flow) |
- High-Resolution Optical Flow from 1D Attention and Correlation
- Learning to Estimate Hidden Motions with Global Motion Aggregation
-
Data Synthesis
- Learning Optical Flow from Still Images
- AutoFlow: Learning a Better Training Set for Optical Flow - research/opticalflow-autoflow) 
- Optical Flow Dataset Synthesis from Unpaired Images
- AutoFlow: Learning a Better Training Set for Optical Flow - research/opticalflow-autoflow) 
- Learning Optical Flow from Still Images
- Optical Flow Dataset Synthesis from Unpaired Images
-
Unsupervised Models
- Optical Flow Training under Limited Label Budget via Active Learning - flow-active-learning-release](https://github.com/duke-vision/optical-flow-active-learning-release) 
- SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping - research/google-research/tree/master/smurf) GoogleResearch
- UPFlow: Upsampling Pyramid for Unsupervised Optical Flow Learning
- Optical Flow Training under Limited Label Budget via Active Learning - flow-active-learning-release](https://github.com/duke-vision/optical-flow-active-learning-release) 
- SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping - research/google-research/tree/master/smurf) GoogleResearch
- UPFlow: Upsampling Pyramid for Unsupervised Optical Flow Learning
- OccInpFlow: Occlusion-Inpainting Optical Flow Estimation by Unsupervised Learning
- What Matters in Unsupervised Optical Flow - research/google-research/tree/master/uflow) GoogleResearch
- Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation
- Flow2Stereo: Effective Self-Supervised Learning of Optical Flow and Stereo Matching
- OccInpFlow: Occlusion-Inpainting Optical Flow Estimation by Unsupervised Learning
- What Matters in Unsupervised Optical Flow - research/google-research/tree/master/uflow) GoogleResearch
- Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation
- Flow2Stereo: Effective Self-Supervised Learning of Optical Flow and Stereo Matching
-
Multi-Frame Supervised Models
- Local All-Pair Correspondence for Point Tracking
- FlowTrack: Revisiting Optical Flow for Long-Range Dense Tracking
- Dense Optical Tracking: Connecting the Dots
- Local All-Pair Correspondence for Point Tracking
- FlowTrack: Revisiting Optical Flow for Long-Range Dense Tracking
- Dense Optical Tracking: Connecting the Dots
- Tracking Everything Everywhere All at Once
- Tracking Everything Everywhere All at Once
- AccFlow: Backward Accumulation for Long-Range Optical Flow
- VideoFlow: Exploiting Temporal Cues for Multi-frame Optical Flow Estimation
- Particle Video Revisited: Tracking Through Occlusions Using Point Trajectories
- AccFlow: Backward Accumulation for Long-Range Optical Flow
- VideoFlow: Exploiting Temporal Cues for Multi-frame Optical Flow Estimation
- Particle Video Revisited: Tracking Through Occlusions Using Point Trajectories
-
Semi-Supervised Models
-
Joint Learning
- Unifying Flow, Stereo and Depth Estimation
- Unifying Flow, Stereo and Depth Estimation
- EffiScene: Efficient Per-Pixel Rigidity Inference for Unsupervised Joint Learning of Optical Flow, Depth, Camera Pose and Motion Segmentation
- Feature-Level Collaboration: Joint Unsupervised Learning of Optical Flow, Stereo Depth and Camera Motion
- EffiScene: Efficient Per-Pixel Rigidity Inference for Unsupervised Joint Learning of Optical Flow, Depth, Camera Pose and Motion Segmentation
- Feature-Level Collaboration: Joint Unsupervised Learning of Optical Flow, Stereo Depth and Camera Motion
-
Special Scene
- Unsupervised Cumulative Domain Adaptation for Foggy Scene Optical Flow - Flow](https://github.com/hyzhouboy/UCDA-Flow) 
- Unsupervised Cumulative Domain Adaptation for Foggy Scene Optical Flow - Flow](https://github.com/hyzhouboy/UCDA-Flow) 
- Deep 360∘ Optical Flow Estimation Based on Multi-Projection Fusion
- Optical flow estimation from a single motion-blurred image
- Optical Flow in Dense Foggy Scenes using Semi-Supervised Learning
- Optical Flow in the Dark - Flow-in-the-Dark](https://github.com/mf-zhang/Optical-Flow-in-the-Dark) 
- Deep 360∘ Optical Flow Estimation Based on Multi-Projection Fusion
- Optical flow estimation from a single motion-blurred image
- Optical Flow in Dense Foggy Scenes using Semi-Supervised Learning
- Optical Flow in the Dark - Flow-in-the-Dark](https://github.com/mf-zhang/Optical-Flow-in-the-Dark) 
-
Special Device
- event-based_vision_resources - rpg/event-based_vision_resources#optical-flow-estimation)
- Learning Optical Flow from Event Camera with Rendered Dataset
- Secrets of Event-Based Optical Flow - rip/event_based_optical_flow) 
- GyroFlow: Gyroscope-Guided Unsupervised Optical Flow Learning - research/GyroFlow) 
- event-based_vision_resources - rpg/event-based_vision_resources#optical-flow-estimation)
- Learning Optical Flow from Event Camera with Rendered Dataset
- Secrets of Event-Based Optical Flow - rip/event_based_optical_flow) 
-
-
Applications
-
Low Level Vision
- Deep Burst Super-Resolution - burst-sr](https://github.com/goutamgmb/deep-burst-sr) 
- Efficient Dynamic Scene Deblurring Using Spatially Variant Deconvolution Network With Optical Flow Guided Training
- Deep video super-resolution using HR optical flow estimation - VSR](https://github.com/The-Learning-And-Vision-Atelier-LAVA/SOF-VSR) 
- Deep Reparametrization of Multi-Frame Super-Resolution and Denoising - rep](https://github.com/goutamgmb/deep-rep) 
- Deep Burst Super-Resolution - burst-sr](https://github.com/goutamgmb/deep-burst-sr) 
- Efficient Dynamic Scene Deblurring Using Spatially Variant Deconvolution Network With Optical Flow Guided Training
- Deep video super-resolution using HR optical flow estimation - VSR](https://github.com/The-Learning-And-Vision-Atelier-LAVA/SOF-VSR) 
- Deep Reparametrization of Multi-Frame Super-Resolution and Denoising - rep](https://github.com/goutamgmb/deep-rep) 
-
Video Synthesis/Generation
- Neighbor correspondence matching for flow-based video frame synthesis
- Clearer Frames, Anytime: Resolving Velocity Ambiguity in Video Frame Interpolation - Clearer](https://github.com/zzh-tech/InterpAny-Clearer) 
- MoVideo: Motion-Aware Video Generation with Diffusion Models
- FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis
- Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution - research/WACV2024-SAFA) 
- A Dynamic Multi-Scale Voxel Flow Network for Video Prediction - research/CVPR2023-DMVFN) 
- Conditional Image-to-Video Generation with Latent Flow Diffusion Models
- A Unified Pyramid Recurrent Network for Video Frame Interpolation - Net](https://github.com/srcn-ivl/UPR-Net) 
- Extracting Motion and Appearance via Inter-Frame Attention for Efficient Video Frame Interpolation - VFI](https://github.com/MCG-NJU/EMA-VFI) 
- Frame Interpolation for Dynamic Scenes with Implicit Flow Encoding
- Clearer Frames, Anytime: Resolving Velocity Ambiguity in Video Frame Interpolation - Clearer](https://github.com/zzh-tech/InterpAny-Clearer) 
- MoVideo: Motion-Aware Video Generation with Diffusion Models
- FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis
- Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution - research/WACV2024-SAFA) 
- A Dynamic Multi-Scale Voxel Flow Network for Video Prediction - research/CVPR2023-DMVFN) 
- Conditional Image-to-Video Generation with Latent Flow Diffusion Models
- A Unified Pyramid Recurrent Network for Video Frame Interpolation - Net](https://github.com/srcn-ivl/UPR-Net) 
- Extracting Motion and Appearance via Inter-Frame Attention for Efficient Video Frame Interpolation - VFI](https://github.com/MCG-NJU/EMA-VFI) 
- Frame Interpolation for Dynamic Scenes with Implicit Flow Encoding
- Neighbor correspondence matching for flow-based video frame synthesis
- VideoINR: Learning Video Implicit Neural Representation for Continuous Space-Time Super-Resolution - AI-Research/VideoINR-Continuous-Space-Time-Super-Resolution) 
- IFRNet: Intermediate Feature Refine Network for Efficient Frame Interpolation
- Deep Animation Video Interpolation in the Wild
- Softmax Splatting for Video Frame Interpolation - splatting](https://github.com/sniklaus/softmax-splatting) 
- Adaptive Collaboration of Flows for Video Frame Interpolation - pytorch](https://github.com/HyeongminLEE/AdaCoF-pytorch) 
- FeatureFlow: Robust Video Interpolation via Structure-to-Texture Generation - BF/FeatureFlow) 
- VideoINR: Learning Video Implicit Neural Representation for Continuous Space-Time Super-Resolution - AI-Research/VideoINR-Continuous-Space-Time-Super-Resolution) 
- IFRNet: Intermediate Feature Refine Network for Efficient Frame Interpolation
- Deep Animation Video Interpolation in the Wild
- Adaptive Collaboration of Flows for Video Frame Interpolation - pytorch](https://github.com/HyeongminLEE/AdaCoF-pytorch) 
- FeatureFlow: Robust Video Interpolation via Structure-to-Texture Generation - BF/FeatureFlow) 
- Real-Time Intermediate Flow Estimation for Video Frame Interpolation - RIFE) 
- Real-Time Intermediate Flow Estimation for Video Frame Interpolation - RIFE) 
- Neural Frame Interpolation for Rendered Content
-
Video Inpainting
-
Video Stabilization
-
Stereo and SLAM
- RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching - Stereo](https://github.com/princeton-vl/RAFT-Stereo) 
- VOLDOR: Visual Odometry From Log-Logistic Dense Optical Flow Residuals
- RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching - Stereo](https://github.com/princeton-vl/RAFT-Stereo) 
- VOLDOR: Visual Odometry From Log-Logistic Dense Optical Flow Residuals
-
-
Scene Flow
-
Special Device
- RAFT-3D: Scene Flow Using Rigid-Motion Embeddings
- Just Go With the Flow: Self-Supervised Scene Flow Estimation - Go-with-the-Flow-Self-Supervised-Scene-Flow-Estimation](https://github.com/HimangiM/Just-Go-with-the-Flow-Self-Supervised-Scene-Flow-Estimation) 
- Just Go With the Flow: Self-Supervised Scene Flow Estimation - Go-with-the-Flow-Self-Supervised-Scene-Flow-Estimation](https://github.com/HimangiM/Just-Go-with-the-Flow-Self-Supervised-Scene-Flow-Estimation) 
- RAFT-3D: Scene Flow Using Rigid-Motion Embeddings
- Learning to Segment Rigid Motions from Two Frames - y/rigidmask)
- Upgrading Optical Flow to 3D Scene Flow through Optical Expansion - y/expansion) 
- Self-Supervised Monocular Scene Flow Estimation - mono-sf](https://github.com/visinf/self-mono-sf) 
- Learning to Segment Rigid Motions from Two Frames - y/rigidmask)
- Upgrading Optical Flow to 3D Scene Flow through Optical Expansion - y/expansion) 
- Self-Supervised Monocular Scene Flow Estimation - mono-sf](https://github.com/visinf/self-mono-sf) 
-
Programming Languages
Sub Categories
Others
45
Classical Estimation Methods
40
Supervised Models
34
Video Synthesis/Generation
34
Special Device
17
Unsupervised Models
14
Multi-Frame Supervised Models
14
Special Scene
10
Low Level Vision
8
Data Synthesis
6
Joint Learning
6
Video Inpainting
4
Stereo and SLAM
4
Semi-Supervised Models
2
Video Stabilization
2
Keywords