Abstract:The degradation of image resolution reduces the detection performance in aerial imagery because it generates a large number of small objects, and accurately detecting these small objects remains a challenge. Existing methods mostly use a superresolution (SR) model to first obtain the SR image of the low-resolution degraded image ( ) and then use this image as the input of the object detection (OD) network to solve this problem. However, this architecture that involves executing a complex SR network before the detector is time-consuming and makes it hard to achieve real-time model inference. To address this challenge, we propose a simple and effective rotated small OD method, named end-to-end superresolution enhanced real-time rotated object detector (ESRTMDet). First, we design a lightweight embedded feature map superresolution module (ESRM) embedded in the detection model to enhance and amplify the backbone output features, making the detection heads detect small objects more easily. Furthermore, we train a parallel SR network branch (PSRB) simultaneously that uses the backbone feature to restore a high-resolution image. Through our proposed feature alignment loss and feature affinity layer, our PSRB effectively guides the feature map enhancement of ESRM. Finally, through end-to-end joint optimization of the detector and PSRB, the detection performance on is significantly improved. Extensive experiments over DOTA and UCAS-AOD demonstrate that our method can achieve state-of-the-art results. In addition, we discard our PSRB and use as the input during inference, reducing the inference time-consuming of our model. Therefore, our ESRTMDet-X not only achieves 77.11% mean of average precision on - he degraded DOTA dataset, but also achieves an amazing inference speed of 337 FPS, thus obtaining the best speed–accuracy tradeoff.

Object detection on low-resolution images with two-stage enhancement

TIENet: task-oriented image enhancement network for degraded object detection

ESRTMDet: An End-to-End Super-Resolution Enhanced Real-Time Rotated Object Detector for Degraded Aerial Images

RestoreDet: Degradation Equivariant Representation for Object Detection in Low Resolution Images

An Improved DETR Based on Angle Denoising and Oriented Boxes Refinement for Remote Sensing Object Detection

Low-quality Image Object Detection Based on Reinforcement Learning Adaptive Enhancement

Dynamic Low-Light Image Enhancement for Object Detection Via End-to-End Training.

Small-Object Detection in Remote Sensing Images With Super-Resolution Perception

HTD: Heterogeneous Task Decoupling for Two-Stage Object Detection

Object Detection for Remote Sensing Based on the Enhanced YOLOv8 With WBiFPN

Task-Aligned Oriented Object Detection in Remote Sensing Images

LEDet: A Single-Shot Real-Time Object Detector Based on Low-Light Image Enhancement

Image Processing: Facilitating Retinanet for Detecting Small Objects

Task-Balanced Distillation for Object Detection

Single-Shot Refinement Neural Network for Object Detection

Degradation Type-Aware Image Restoration for Effective Object Detection in Adverse Weather

Joint Anchor-Feature Refinement for Real-Time Accurate Object Detection in Images and Videos

Comprehensive Feature Enhancement Module For Single-Shot Object Detector

Multi-stage Enhancement Network for Tiny Object Detection in Remote Sensing Images

RESC: REfine the SCore with Adaptive Transformer Head for End-to-end Object Detection

Improving Oriented Object Detection by Scene Classification and Task-Aligned Focal Loss