The 8-Point Algorithm as an Inductive Bias for Relative Pose Prediction by ViTs
Chris Rockwell, Justin Johnson, David F. Fouhey
Abstract
We present a simple baseline for directly estimating the relative pose (rotation and translation, including scale) between two images. Deep methods have recently shown strong progress but often require complex or multi-stage architectures. We show that a handful of modifications can be applied to a Vision Transformer (ViT) to bring its computations close to the Eight-Point Algorithm. This inductive bias enables a simple method to be competitive in multiple settings, often substantially improving over the state of the art with strong performance gains in limited data regimes.
Topics & Concepts
Computer scienceComputationInductive biasRotation (mathematics)AlgorithmTransformerArtificial intelligenceSimple (philosophy)Point (geometry)Translation (biology)PoseMachine learningMathematicsTask (project management)Multi-task learningVoltageEngineeringGeometryMessenger RNAEpistemologyPhilosophyGeneBiochemistryChemistrySystems engineeringElectrical engineeringRobotics and Sensor-Based LocalizationAdvanced Image and Video Retrieval TechniquesAdvanced Vision and Imaging