SSPNet: Scale Selection Pyramid Network for Tiny Person Detection from UAV Images

TL;DR

This paper introduces SSPNet, a novel scale selection pyramid network designed to improve tiny person detection in UAV images by addressing scale variability and gradient inconsistency issues, achieving superior performance on benchmarks.

Contribution

The paper proposes SSPNet with three modules (CAM, SEM, SSM) and a WNS strategy, enhancing tiny object detection by better scale focus and gradient consistency.

Findings

Outperforms state-of-the-art detectors on TinyPerson benchmark

Effectively highlights scale-specific features for tiny objects

Improves detection accuracy in UAV imagery

Abstract

With the increasing demand for search and rescue, it is highly demanded to detect objects of interest in large-scale images captured by Unmanned Aerial Vehicles (UAVs), which is quite challenging due to extremely small scales of objects. Most existing methods employed Feature Pyramid Network (FPN) to enrich shallow layers' features by combing deep layers' contextual features. However, under the limitation of the inconsistency in gradient computation across different layers, the shallow layers in FPN are not fully exploited to detect tiny objects. In this paper, we propose a Scale Selection Pyramid network (SSPNet) for tiny person detection, which consists of three components: Context Attention Module (CAM), Scale Enhancement Module (SEM), and Scale Selection Module (SSM). CAM takes account of context information to produce hierarchical attention heatmaps. SEM highlights features of specific scales at different layers, leading the detector to focus on objects of specific scales instead of vast backgrounds. SSM exploits adjacent layers' relationships to fulfill suitable feature sharing between deep layers and shallow layers, thereby avoiding the inconsistency in gradient computation across different layers. Besides, we propose a Weighted Negative Sampling (WNS) strategy to guide the detector to select more representative samples. Experiments on the TinyPerson benchmark show that our method outperforms other state-of-the-art (SOTA) detectors.