Few-Shot Batch Incremental Road Object Detection via Detector Fusion

TL;DR

This paper introduces DualFusion, a detector fusion method for batch incremental few-shot road object detection, achieving state-of-the-art results on IDD and COCO datasets by effectively detecting rare objects with minimal data.

Contribution

The paper proposes DualFusion, a novel detector fusion approach that enhances incremental few-shot detection performance without degrading existing class detection accuracy.

Findings

Achieved highest mAP50 scores on IDD base and overall classes.

Surpassed state-of-the-art novel class AP scores on COCO by over 6.6 times.

Demonstrated effective detection of rare objects with limited data.

Abstract

Incremental few-shot learning has emerged as a new and challenging area in deep learning, whose objective is to train deep learning models using very few samples of new class data, and none of the old class data. In this work we tackle the problem of batch incremental few-shot road object detection using data from the India Driving Dataset (IDD). Our approach, DualFusion, combines object detectors in a manner that allows us to learn to detect rare objects with very limited data, all without severely degrading the performance of the detector on the abundant classes. In the IDD OpenSet incremental few-shot detection task, we achieve a mAP50 score of 40.0 on the base classes and an overall mAP50 score of 38.8, both of which are the highest to date. In the COCO batch incremental few-shot detection task, we achieve a novel AP score of 9.9, surpassing the state-of-the-art novel class performance on the same by over 6.6 times.