SOTAVerified

Motion Estimation

Motion Estimation is used to determine the block-wise or pixel-wise motion vectors between two frames.

Source: MEMC-Net: Motion Estimation and Motion Compensation Driven Neural Network for Video Interpolation and Enhancement

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 151–160 of 741 papers

TitleStatusHype
Instance-wise Depth and Motion Learning from Monocular VideosCode1
Restoration of Non-rigidly Distorted Underwater Images using a Combination of Compressive Sensing and Local Polynomial Image RepresentationsCode1
Unsupervised Learning of Depth and Ego-Motion from Cylindrical Panoramic VideoCode1
Taking a Deeper Look at the Inverse Compositional AlgorithmCode1
Joint Unsupervised Learning of Optical Flow and Depth by Watching Stereo VideosCode1
Digging Into Self-Supervised Monocular Depth EstimationCode1
DeepPhys: Video-Based Physiological Measurement Using Convolutional Attention NetworksCode1
Video Enhancement with Task-Oriented FlowCode1
DINO-VO: A Feature-based Visual Odometry Leveraging a Visual Foundation Model—0
Uncertainty-Driven Radar-Inertial Fusion for Instantaneous 3D Ego-Velocity Estimation—0
Show:102550
← PrevPage 16 of 75Next →

No leaderboard results yet.