Improved YOLOX Remote Sensing Image Object Detection Algorithm | Wuhan University Journal of Natural Sciences

Open Access

Table 5

Comparison of ablation experiments of each module of the improved algorithm

YOLOX-S	α-IoU+Adamx	ECA (between Backbone and Neck)	ECA (between Neck and Head)	mAP/%	mAP50:95/%
√	×	×	×	94.2	61.85
√	√	×	×	94.5	62.70
√	√	√	×	95.0	63.05
√	√	√	√	95.5	62.86

"√" indicates the structure is included in the model, and "×" indicates that the structure is not included in the model. The bold font in the table is the maximum value

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.