Scholarlink

With the widespread application of UAV surveillance in social security, Person re-identification (ReID) has naturally extended its scope to the field of unmanned aerial vehicles. However, traditional ground-based person re-identification (ReID) tasks typically rely on surveillance cameras with fixed positions and shooting angles. Under such conditions, pedestrian images exhibit limited resolution variation and implicitly assume an upright body posture perpendicular to the ground. However, in aerial scenarios, image resolution varies significantly with UAV flight altitude, while pedestrian poses demonstrate random translation and rotation phenomena. To address these challenges, this paper proposes an Adaptive Positional Encoding and Multi-scale Self-attention Transformer (APE-MSAT). The Adaptive Positional Encoding module leverages convolutional invariance to preserve robust neighborhood information against random pose variations, while the Multi-scale Self-attention Transformer employs progressively scaled feature maps to handle resolution variations. Our method achieves competitive results on aerial ReID datasets (PRAI-1581 and UAV-Human) as well as the ground-based Market-1501 benchmark.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Adaptive Positional Encoding and Multi-scale Self-attention Transformer for Aerial Person Re-identification

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Zhizhi Lu,
  • Jianhuang Lai

摘要

With the widespread application of UAV surveillance in social security, Person re-identification (ReID) has naturally extended its scope to the field of unmanned aerial vehicles. However, traditional ground-based person re-identification (ReID) tasks typically rely on surveillance cameras with fixed positions and shooting angles. Under such conditions, pedestrian images exhibit limited resolution variation and implicitly assume an upright body posture perpendicular to the ground. However, in aerial scenarios, image resolution varies significantly with UAV flight altitude, while pedestrian poses demonstrate random translation and rotation phenomena. To address these challenges, this paper proposes an Adaptive Positional Encoding and Multi-scale Self-attention Transformer (APE-MSAT). The Adaptive Positional Encoding module leverages convolutional invariance to preserve robust neighborhood information against random pose variations, while the Multi-scale Self-attention Transformer employs progressively scaled feature maps to handle resolution variations. Our method achieves competitive results on aerial ReID datasets (PRAI-1581 and UAV-Human) as well as the ground-based Market-1501 benchmark.