New FGAA-FPN enhances oriented object detection in high-res imagery

Researchers have developed FGAA-FPN, a novel Feature Pyramid Network designed to improve oriented object detection in high-resolution imagery. This network incorporates a Foreground-Guided Feature Modulation module to enhance object regions and suppress background noise, alongside an Angle-Aware Multi-Head Attention module that leverages orientation priors for better feature discrimination. Experiments on the DOTA datasets show FGAA-FPN achieving state-of-the-art performance, with mAP scores of 75.5% on DOTA-v1.0 and 68.3% on DOTA-v1.5. AI

IMPACT This new network could improve accuracy in applications like remote sensing and surveillance by better identifying objects at various orientations.

RANK_REASON The cluster contains a research paper detailing a new technical approach for a specific computer vision task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• DOTA-v1.0
• DOTA-v1.5
• Feature Pyramid Networks for Object Detection
• FGAA-FPN
• Jialin Mao
• Oriented Object Detection

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →