MAFE R-CNN: Selecting More Samples to Learn Category-aware Features for Small Object Detection

TL;DR

This paper introduces MAFE R-CNN, a novel approach for small object detection that improves feature learning and sample selection through multi-clue strategies and category-aware feature enhancement, leading to better detection performance.

Contribution

The paper proposes the Multi-Clue Sample Selection and Category-aware Feature Enhancement mechanisms to improve small object detection by better sample diversity and feature representation.

Findings

Enhanced detection accuracy on SODA dataset

Effective sample selection balancing object sizes

Improved feature representation for small objects

Abstract

Small object detection in intricate environments has consistently represented a major challenge in the field of object detection. In this paper, we identify that this difficulty stems from the detectors' inability to effectively learn discriminative features for objects of small size, compounded by the complexity of selecting high-quality small object samples during training, which motivates the proposal of the Multi-Clue Assignment and Feature Enhancement R-CNN.Specifically, MAFE R-CNN integrates two pivotal components.The first is the Multi-Clue Sample Selection (MCSS) strategy, in which the Intersection over Union (IoU) distance, predicted category confidence, and ground truth region sizes are leveraged as informative clues in the sample selection process. This methodology facilitates the selection of diverse positive samples and ensures a balanced distribution of object sizes during training, thereby promoting effective model learning.The second is the Category-aware Feature Enhancement Mechanism (CFEM), where we propose a simple yet effective category-aware memory module to explore the relationships among object features. Subsequently, we enhance the object feature representation by facilitating the interaction between category-aware features and candidate box features.Comprehensive experiments conducted on the large-scale small object dataset SODA validate the effectiveness of the proposed method. The code will be made publicly available.