Practical Insights into Semi-Supervised Object Detection Approaches

TL;DR

This paper compares three state-of-the-art semi-supervised object detection methods across multiple datasets to understand their performance trade-offs in low-data scenarios, providing practical insights for real-world applications.

Contribution

It offers a comprehensive experimental comparison of MixPL, Semi-DETR, and Consistent-Teacher on standard and custom datasets, highlighting their strengths and limitations.

Findings

Performance varies significantly with the number of labeled images.

Different methods have trade-offs between accuracy, model size, and latency.

Insights guide choosing suitable SSOD approaches for low-data settings.

Abstract

Learning in data-scarce settings has recently gained significant attention in the research community. Semi-supervised object detection(SSOD) aims to improve detection performance by leveraging a large number of unlabeled images alongside a limited number of labeled images(a.k.a.,few-shot learning). In this paper, we present a comprehensive comparison of three state-of-the-art SSOD approaches, including MixPL, Semi-DETR and Consistent-Teacher, with the goal of understanding how performance varies with the number of labeled images. We conduct experiments using the MS-COCO and Pascal VOC datasets, two popular object detection benchmarks which allow for standardized evaluation. In addition, we evaluate the SSOD approaches on a custom Beetle dataset which enables us to gain insights into their performance on specialized datasets with a smaller number of object categories. Our findings highlight the trade-offs between accuracy, model size, and latency, providing insights into which methods are best suited for low-data regimes.