Belangrijkste concepten
Improving object detection benchmarking by refining COCO annotations leads to better model performance.
Statistieken
Due to the prevalence of COCO, we choose to correct errors to maintain continuity with prior research.
Models that predict visually sharper masks score higher on COCO-ReM.
Models trained using COCO-ReM converge faster and score higher than those trained using COCO-2017.
Citaten
"Models that predict visually sharper masks score higher on COCO-ReM."
"Data quality plays a significant role in enhancing object detector capabilities."