Abstract:Most anchor-free methods perform object detection using dense recommendation, which assumes that one point can simultaneously conduct accurate category prediction and regression estimation. However, due to different task drivers, valid features for classification and regression may locate at distinct areas in the training phase. This problem is called feature misalignment. To solve it, we propose a new feature alignment method based on anchor-free object detector. Firstly, a global receptive field adaptor (G-RFA) is designed by incorporating the feature pyramid networks (FPN) with the global attention mechanism, and forward features are further fine-tuned with a deformable-subnet (De-Subnet) to remove the influence of redundant contextual information. Then, a new feature filter strategy with a misalignment score is proposed to guide the network to focus on sampling points with aligned features. In addition, we establish mutually independent multi-layer quality distributions to model the priori information of an object on different FPN levels. Equipped with our method, the classification and regression features are aligned, and the generated foreground weight map converges to the centers of classification and regression heatmaps. Experimental results show that without bells and whistles, our method achieves 49.3% AP on MS COCO test-dev under the default <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2\times $ </tex-math></inline-formula> training schedule, outperforming related methods. Besides, experiments on PASCAL VOC demonstrate the generalization ability of our method. Code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/GFENGG/featurealign</uri> .

Label Assignment for Aligning Features in Anchor-free Models

Feature Alignment in Anchor-Free Object Detection

Automatic label assignment object detection mehtod on only one feature map

Branch Aware Assignment for Object Detection

An Anchor-Free Network with Density Map and Attention Mechanism for Multiscale Object Detection in Aerial Images

Instance-based Dynamic Label Assignment for Object Detection

Sparse Label Assignment for Oriented Object Detection in Aerial Images

Pondering About Task Spatial Misalignment: Classification-Localization Equilibrated Object Detection

Balance-Oriented Focal Loss with Linear Scheduling for Anchor Free Object Detection

PosNeg-Balanced Anchors with Aligned Features for Single-Shot Object Detection

Learning from Noisy Anchors for One-Stage Object Detection.

Align Deep Features for Oriented Object Detection

AlignYOLO: A feature-aligned network for object detection

A dynamic label assignment strategy for one-stage detectors

NLFA: A Non Local Fusion Alignment Module for Multi-Scale Feature in Object Detection

Cross-domain Object Detection by Local to Global Object-Aware Feature Alignment

Level-wise Dynamic Label Assignment for Oriented Object Detection.

Dynamic Anchor Feature Selection for Single-Shot Object Detection

Object Detection for Aerial Images With Feature Enhancement and Soft Label Assignment

Multi-task Feature-Aligned Head in One-Stage Object Detection

Anchor Free Network for Multi-Scale Face Detection