Towards Balanced Learning for Instance Recognition

TL;DR

This paper introduces Libra R-CNN, a framework that addresses training imbalance at sample, feature, and objective levels to improve instance recognition performance in deep learning detectors.

Contribution

It proposes a novel balanced learning framework integrating IoU-balanced sampling, balanced feature pyramid, and objective re-weighting, enhancing detector training.

Findings

Improved detection accuracy on MS COCO, LVIS, and Pascal VOC datasets.

Effective reduction of training imbalance at multiple levels.

Significant performance gains over baseline detectors.

Abstract

Instance recognition is rapidly advanced along with the developments of various deep convolutional neural networks. Compared to the architectures of networks, the training process, which is also crucial to the success of detectors, has received relatively less attention. In this work, we carefully revisit the standard training practice of detectors, and find that the detection performance is often limited by the imbalance during the training process, which generally consists in three levels - sample level, feature level, and objective level. To mitigate the adverse effects caused thereby, we propose Libra R-CNN, a simple yet effective framework towards balanced learning for instance recognition. It integrates IoU-balanced sampling, balanced feature pyramid, and objective re-weighting, respectively for reducing the imbalance at sample, feature, and objective level. Extensive experiments conducted on MS COCO, LVIS and Pascal VOC datasets prove the effectiveness of the overall balanced design.