English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper presents a novel architecture for simultaneous estimation ofhighly accurate optical flows and rigid scene transformations for difficultscenarios where the brightness assumption is violated by strong shadingchanges. In the case of rotating objects or moving light sources, such as thoseencountered for driving cars in the dark, the scene appearance often changessignificantly from one view to the next. Unfortunately, standard methods forcalculating optical flows or poses are based on the expectation that theappearance of features in the scene remain constant between views. Thesemethods may fail frequently in the investigated cases. The presented methodfuses texture and geometry information by combining image, vertex and normaldata to compute an illumination-invariant optical flow. By using acoarse-to-fine strategy, globally anchored optical flows are learned, reducingthe impact of erroneous shading-based pseudo-correspondences. Based on thelearned optical flows, a second architecture is proposed that predicts robustrigid transformations from the warped vertex and normal maps. Particularattention is payed to situations with strong rotations, which often cause suchshading changes. Therefore a 3-step procedure is proposed that profitablyexploits correlations between the normals and vertices. The method has beenevaluated on a newly created dataset containing both synthetic and real datawith strong rotations and shading effects. This data represents the typical usecase in 3D reconstruction, where the object often rotates in large stepsbetween the partial reconstructions. Additionally, we apply the method to thewell-known Kitti Odometry dataset. Even if, due to fulfillment of the brighnessassumption, this is not the typical use case of the method, the applicabilityto standard situations and the relation to other methods is thereforeestablished.

INV-Flow2PoseNet: Light-Resistant Rigid Object Pose from Optical Flow of RGB-D Images using Images, Normals and Vertices