Mixup Regularization for Region Proposal based Object Detectors
This work addresses a domain-specific problem for object detection researchers, offering an incremental improvement in regularization techniques.
The paper tackles the challenge of extending mixup regularization to object detection by leveraging anchor-based region mapping, resulting in improved robustness to image alterations and better generalization.
Mixup - a neural network regularization technique based on linear interpolation of labeled sample pairs - has stood out by its capacity to improve model's robustness and generalizability through a surprisingly simple formalism. However, its extension to the field of object detection remains unclear as the interpolation of bounding boxes cannot be naively defined. In this paper, we propose to leverage the inherent region mapping structure of anchors to introduce a mixup-driven training regularization for region proposal based object detectors. The proposed method is benchmarked on standard datasets with challenging detection settings. Our experiments show an enhanced robustness to image alterations along with an ability to decontextualize detections, resulting in an improved generalization power.