DSLA: Dynamic smooth label assignment for efficient anchor-free object detection. (November 2022)

• Record Type: Journal Article
• Authors: Su, Hu; He, Yonghao; Jiang, Rui; Zhang, Jiabin; Zou, Wei; Fan, Bin
• Is Part Of: Pattern recognition. Volume 131(2022)
• Publication Date: 2022-11
• Subjects: Convolutional neural network -- Object detection -- Centerness score -- Intersection-of-union; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2022.108868
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: The inconsistencies of classification and quality estimation are analyzed. Dynamic smooth label assignment is proposed to address the problems. Interval relaxation strategy is proposed and combined with the improved centerness score. The assigned label is smoothed to a continuous value. IoU score is dynamically calculated and coupled with the smooth label, resulting in dynamic smooth label. DSLA is applied to popular anchor-free detectors. Comprehensive experiments are carried out on MS COCO to demonstrate the effectiveness. Abstract: Anchor-free detectors basically formulate object detection as dense classification and regression. For popular anchor-free detectors, it is common to introduce an individual prediction branch to estimate the quality of localization. The following inconsistencies are observed when we delve into the practices of classification and quality estimation. Firstly, for some adjacent samples which are assigned completely different labels, the trained model would produce similar classification scores. This violates the training objective and leads to performance degradation. Secondly, it is found that detected bounding boxes with higher confidences contrarily have smaller overlaps with the corresponding ground-truth. Accurately localized bounding boxes would be suppressed by less accurate ones in the Non-Maximum Suppression (NMS) procedure. To address the inconsistency problems, the Dynamic Smooth Label Assignment (DSLA) method is proposed. Based on the concept of centerness originally developed in FCOS, a smooth assignment strategy is proposed. The label is smoothed to a continuous value in [ 0, 1 ] to make a steady transition between positive and negative samples. Intersection-of-Union (IoU) is predicted dynamically during training and is coupled with the smoothed label. The dynamic smooth label is assigned to supervise the classification branch. Under such supervision, quality estimation branch is naturally merged into the classification branch, which simplifies the architecture of anchor-free detector. Comprehensive experiments are conducted on the MS COCO benchmark. It is demonstrated that, DSLA can significantly boost the detection accuracy by alleviating the above inconsistencies for anchor-free detectors. Our codes are released at https://github.com/YonghaoHe/DSLA .