What question did this study set out to answer?

To develop Light-YOLOv7+, a lightweight target detection algorithm that enhances multi-scale feature representation and inference speed under strict computational constraints in remote sensing imagery.

August 22, 2026International Journal of Computational Intelligence Systems0 citationsOpen Access

Lightweight Remote Sensing Image Target Detection Algorithm Based on Improved YOLOv7

CYChengyu YangNortheastern University

Key Points

To develop Light-YOLOv7+, a lightweight target detection algorithm that enhances multi-scale feature representation and inference speed under strict computational constraints in remote sensing imagery.
Integrated a Quantized Efficient Feature Fusion Module (QEFFM) and a Spatial–Temporal Dual Attention Module (STDAM) to aggregate multi-scale features and adaptively reweight spatial-contextual responses.
Incorporated an adjustable scale feature pyramid, dynamic anchor generation, and an ultra-lightweight decoding head to lower floating-point operations (FLOPs) and latency.
Evaluated detection accuracy and efficiency across three remote sensing benchmark datasets: NWPU VHR-10, DOTA, and RSOD.
Achieved a mean average precision (mAP) of 0.878 on the NWPU VHR-10 dataset, an increase of +0.062 over the baseline YOLOv7.
Attained a recall of 0.847 on the DOTA dataset while concurrently reducing model complexity, FLOPs, and inference latency.

Abstract

High detection accuracy under computing restrictions is a major concern as remote sensing imagery grows exponentially. Light-YOLOv7+, a lightweight target detection system, improves multi-scale representation and inference performance by coordinated architectural modifications. The proposed design uses a Quantized Efficient Feature Fusion Module (QEFFM) and a Spatial–Temporal Dual Attention Module (STDAM) to boost discriminative features across pyramid levels by aggregating features at low cost and reducing redundancy. STDAM adaptively reweights spatial–contextual responses. In addition, an adjustable scale feature pyramid, dynamic anchor creation, and an ultra-lightweight decoding head reduce computing cost while maintaining fine-grained localization. Testing on NWPU VHR-10, DOTA, and RSOD shows continuous performance increases under model capacity constraints. On NWPU VHR-10, Light-YOLOv7 + has a mAP of 0.878 (+ 0.062 over YOLOv7) and a DOTA recall (0.847). While improving accuracy, the model also reduces complexity, FLOPs, and inference latency, making it suitable for resource-limited deployments. The framework is suitable for large-scale remote sensing applications like environmental monitoring and catastrophe assessment because lightweight feature fusion and attention modulation balance efficiency and detection accuracy.

KI fragen

Bookmark

View Full Paper