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

3D object detection with point clouds and images plays an important role inperception tasks such as autonomous driving. Current methods show greatperformance on detection and pose estimation of standard-shaped vehicles butlack behind on more complex shapes as e.g. semi-trailer truck combinations.Determining the shape and motion of those special vehicles accurately iscrucial in yard operation and maneuvering and industrial automationapplications. This work introduces several new methods to improve and measurethe performance for such classes. State-of-the-art methods are based onpredefined anchor grids or heatmaps for ground truth targets. However, theunderlying representations do not take the shape of different sized objectsinto account. Our main contribution, AdaptiveShape, uses shape aware anchordistributions and heatmaps to improve the detection capabilities. For largevehicles we achieve +10.9% AP in comparison to current shape agnostic methods.Furthermore we introduce a new fast LiDAR-camera fusion. It is based on 2Dbounding box camera detections which are available in many processingpipelines. This fusion method does not rely on perfectly calibrated ortemporally synchronized systems and is therefore applicable to a broad range ofrobotic applications. We extend a standard point pillar network to account fortemporal data and improve learning of complex object movements. In addition weextended a ground truth augmentation to use grouped object pairs to furtherimprove truck AP by +2.2% compared to conventional augmentation.

AdaptiveShape: Solving Shape Variability for 3D Object Detection with Geometry Aware Anchor Distributions