Object detection with transformer on many domains by attention to past predictions with Cascade-DETR
Object detection with transformer on many domains by attention to past predictions with Cascade-DETR
Cascade-DETR: Delving into High-Quality Universal Object Detection
arXiv paper abstract https://arxiv.org/abs/2307.11035
arXiv PDF paper https://arxiv.org/pdf/2307.11035.pdf
… While dominating on the COCO benchmark, recent Transformer-based detection methods are not competitive in diverse domains.
Moreover, these methods still struggle to very accurately estimate the object bounding boxes in complex environments.
… introduce Cascade-DETR for high-quality universal object detection.
… jointly tackle the generalization to diverse domains and localization accuracy by proposing the Cascade Attention layer, which explicitly integrates object-centric information into the detection decoder by limiting the attention to the previous box prediction.
… Instead of relying on classification scores, … predict the expected IoU of the query, leading to substantially more well-calibrated confidences.
… While also advancing the state-of-the-art on COCO, Cascade-DETR substantially improves DETR-based detectors on all datasets in UDB10, even by over 10 mAP in some cases …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
