AP-Loss for Accurate One-Stage Object Detection

TL;DR

This paper introduces AP-loss for one-stage object detection, replacing classification with a ranking task to better handle class imbalance, leading to improved accuracy and state-of-the-art results.

Contribution

It proposes a novel AP-loss based framework with a new optimization algorithm to improve one-stage object detectors by addressing class imbalance.

Findings

Significant performance improvement over existing AP-based methods.

Achieved state-of-the-art results on standard benchmarks.

Demonstrated versatility across different network architectures.

Abstract

One-stage object detectors are trained by optimizing classification-loss and localization-loss simultaneously, with the former suffering much from extreme foreground-background class imbalance issue due to the large number of anchors. This paper alleviates this issue by proposing a novel framework to replace the classification task in one-stage detectors with a ranking task, and adopting the Average-Precision loss (AP-loss) for the ranking problem. Due to its non-differentiability and non-convexity, the AP-loss cannot be optimized directly. For this purpose, we develop a novel optimization algorithm, which seamlessly combines the error-driven update scheme in perceptron learning and backpropagation algorithm in deep networks. We provide in-depth analyses on the good convergence property and computational complexity of the proposed algorithm, both theoretically and empirically. Experimental results demonstrate notable improvement in addressing the imbalance issue in object detection over existing AP-based optimization algorithms. An improved state-of-the-art performance is achieved in one-stage detectors based on AP-loss over detectors using classification-losses on various standard benchmarks. The proposed framework is also highly versatile in accommodating different network architectures. Code is available at https://github.com/cccorn/AP-loss .