Recent advances in small object detection based on deep learning: A review

By Kang Tong, Yiquan Wu, Fei Zhou. Image and Vision Computing, 2020

546 citations recorded by the source corpus.

Abstract Small object detection is a challenging problem in computer vision. It has been widely applied in defense military, transportation, industry, etc. To facilitate in-depth understanding of small object detection, we comprehensively review the existing small object detection methods based on deep learning from five aspects, including multi-scale feature learning, data augmentation, training strategy, context-based detection and GAN-based detection. Then, we thoroughly analyze the performance of some typical small object detection algorithms on popular datasets, such as MS-COCO, PASCAL-VOC. Finally, the possible research directions in the future are pointed out from five perspectives: emerging small object detection datasets and benchmarks, multi-task joint learning and optimization, information transmission, weakly supervised small object detection methods and framework for small object detection task.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Industrial and Manufacturing Engineering

View Paper: https://www.semanticscholar.org/paper/243cfd8867ce54ab3f1291489ac5fb0e7bcff554

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

1. Junwei Han
2. Gong Cheng
3. Bo Li
4. Ke Yan
5. Ziwei Liu
6. Zhenwei Shi
7. Yair Weiss
8. Aharon Azulay

In Computer Vision and Pattern Recognition

1. Junwei Han
2. Gong Cheng
3. Xiwen Yao
4. Ziwei Liu
5. Bo Li
6. Dacheng Tao
7. Luc Van Gool
8. Ming-Hsuan Yang

In Industrial and Manufacturing Engineering

1. Bo Li
2. Gong Cheng

Citations in the corpus

• FFCA-YOLO for Small Object Detection in Remote Sensing Images (2024, background): “…years, the research on small object detection has achieved significant growth due to the rapid development of optical remote sensing technology [1], [2], [3], [4], [5], [6] for applications, such as traffic supervision, search and Yin Zhang, Mu Ye, Guiyi Zhu, and Junhua Yan are with the…”
• OtterHD: A High-Resolution Multi-modality Model (2023, background): “Grasping intricate visual details, particularly those of smaller objects, is crucial for computer vision models to be effectively applied in real-world scenarios such as autonomous driving and robotics [32, 53].”
• Towards Large-Scale Small Object Detection: Survey and Benchmarks (2022, background): “Furthermore, there already exist several reviews paid their attention to small object detection [25], [26], [27], however, they fail to the comprehensiveness and in-depth analysis because only partial reviews on limited areas were conducted.”
• Detecting tiny objects in aerial images: A normalized Wasserstein distance and a new benchmark (2022, unknown): “Tiny Object Detection Strategies Most of the previous small/tiny object detection strategies can be roughly grouped into the following five categories: multi-scale feature learning, context-based detection, data augmentation, designing better training strategy (Tong et al., 2020) and label assignmen”
• A Normalized Gaussian Wasserstein Distance for Tiny Object Detection (2021, unknown): “Most of the previous small/tiny object detection methods can be roughly divided into three categories: multi-scale feature learning, designing better training strategy and GAN-based detection [28].”