Slicing Aided Hyper Inference and Fine-tuning for Small Object Detection

Abstract

Detection of small objects and objects far away in the scene is a major challenge in surveillance applications. Such objects are represented by small number of pixels in the image and lack sufficient details, making them difficult to detect using conventional detectors. In this work, an open-source framework called Slicing Aided Hyper Inference (SAHI) is proposed that provides a generic slicing aided inference and fine-tuning pipeline for small object detection. The proposed technique is generic in the sense that it can be applied on top of any available object detector without any fine-tuning. Experimental evaluations, using object detection baselines on the Visdrone and xView aerial object detection datasets show that the proposed inference method can increase object detection AP by 6.8%, 5.1% and 5.3% for FCOS, VFNet and TOOD detectors, respectively. Moreover, the detection accuracy can be further increased with a slicing aided fine-tuning, resulting in a cumulative increase of 12.7%, 13.4% and 14.5% AP in the same order. Proposed technique has been integrated with Detectron2, MMDetection and YOLOv5 models and it is publicly available at https://github.com/obss/sahi.git .

Presented at ICIP 2022, 5 pages, 4 figures, 2 tables

References in corpus (1)

• YOLOv4: Optimal Speed and Accuracy of Object Detection

Cited by in corpus (13)

• RS-YOLOX: A High Precision Detector for Object Detection in Satellite Remote Sensing Images
• Developing a Hybrid Convolutional Neural Network for Automatic Aphid Counting in Sugar Beet Fields
• MVA2023 Small Object Detection Challenge for Spotting Birds: Dataset, Methods, and Results
• DSORT-MCU: Detecting Small Objects in Real-Time on Microcontroller Units
• VME: A Satellite Imagery Dataset and Benchmark for Detecting Vehicles in the Middle East and Beyond
• Improving Object Detection Quality in Football Through Super-Resolution Techniques
• Rega-Net:Retina Gabor Attention for Deep Convolutional Neural Networks
• Utilizing dataset affinity prediction in object detection to assess training data
• deepNIR: Datasets for generating synthetic NIR images and improved fruit detection system using deep learning techniques
• Collaborative real-time vision-based device for olive oil production monitoring
• DroBoost: An Intelligent Score and Model Boosting Method for Drone Detection
• A Comprehensive Framework for Automated Quality Control in the Automotive Industry
• Maritime Small Object Detection from UAVs using Deep Learning with Altitude-Aware Dynamic Tiling