Region Proposal by Guided Anchoring

By Kai Chen, Shuo Yang, Chen Change Loy, Dahua Lin. 2019

609 citations recorded by the source corpus.

Region anchors are the cornerstone of modern object detection techniques. State-of-the-art detectors mostly rely on a dense anchoring scheme, where anchors are sampled uniformly over the spatial domain with a predefined set of scales and aspect ratios. In this paper, we revisit this foundational stage. Our study shows that it can be done much more effectively and efficiently. Specifically, we present an alternative scheme, named Guided Anchoring, which leverages semantic features to guide the anchoring. The proposed method jointly predicts the locations where the center of objects of interest are likely to exist as well as the scales and aspect ratios at different locations. On top of predicted anchor shapes, we mitigate the feature inconsistency with a feature adaption module. We also study the use of high-quality proposals to improve detection performance. The anchoring scheme can be seamlessly integrated into proposal methods and detectors. With Guided Anchoring, we achieve 9.1% higher recall on MS COCO with 90% fewer anchors than the RPN baseline. We also adopt Guided Anchoring in Fast R-CNN, Faster R-CNN and RetinaNet, respectively improving the detection mAP by 2.2%, 2.7% and 1.2%. Code is available at https://github.com/open-mmlab/mmdetection.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Bridging the Gap Between Anchor-Based and Anchor-Free Detection via Adaptive Training Sample Selection (2019, methodology): “GA-RPN [59] deﬁnes the pixels in the center region of the object as positives to predict the location, width and height of object proposals for Faster R-CNN.”
Learning Open-World Object Proposals without Learning to Classify (2021, methodology): “Guided anchoring (GA-RPN) [6] is an advanced iterative RPN with deformable convolution, and Cascade RPN [5] is the state-of-the-art RPN method with multi-stage refinement.”
Hybrid Task Cascade for Instance Segmentation (2019, methodology): “We finetune trained detectors with the proposals generated by GA-RPN [41], which achieves near 10% higher recall than RPN.”
DetNAS: Backbone Search for Object Detection (2019, background): “In the past few years, many advances in object detection come from the study of "head", such as architecture [13], loss [14, 23], and anchor [28, 25].”
MMDetection: Open MMLab Detection Toolbox and Benchmark (2019, background): “• Guided Anchoring [34]: a new anchoring scheme that predicts sparse and arbitrary-shaped anchors, proposed in 2019.”
YOLOv4: Optimal Speed and Accuracy of Object Detection (2020, methodology): “Region proposal by guided anchoring [82] RetinaNet ResNet-50 - 10.”
Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training (2020, background): “e whole network fully convolutional to improve the speed; and FPN [25] proposes a top-down pathway to combine multi-scale features. Besides, various improvements have been witnessed in recent studies [16,41,23,24,49]. Classication in object detection. Recent researches focus on improving object clas”
Align Deep Features for Oriented Object Detection (2020, background): “Both [26] and [27] achieve feature alignment in a fully convolutional way and enjoy high efficiency.”
Learning Spatial Fusion for Single-Shot Object Detection (2019, methodology): “More-over, a number of studies [38, 45, 36, 14] indicate that the anchor-free pipeline contributes to considerably better performance with simpler designs.”
Soft Anchor-Point Object Detection (2019, methodology): “2 GA-Faster-RCNN [28] R-50 X 9.”
Spatial Feature Calibration and Temporal Fusion for Effective One-stage Video Instance Segmentation (2021, methodology): “Besides, inspired by Guided Anchor [38, 23], SipMask further aligns feature with regressed bounding boxes to improve feature representation for classification and mask coefficients generation.”
Anchor-Free Oriented Proposal Generator for Object Detection (2021, positive): “Cascade RPN [18], following the Guided Anchoring route, runs a cascade structure to adapt anchors gradually.”
CentripetalNet: Pursuing High-Quality Keypoint Pairs for Object Detection (2020, background): “Feature adaption has recently been studied in anchor-based detectors [35] [5], but our work is the first to discuss the”
Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution (2019, methodology): “To alleviate this problem, recent advanced methods [12, 37] rely on deformable convolution [10] to perform feature spatial transformations and expect the learned transformations to align to the changes of anchor geometry.”
Auto-FPN: Automatic Network Architecture Adaptation for Object Detection Beyond Classification (2019, background): “…34.9 FPN [32] ResNet-101 40.7 Relation Network [21] ResNet-101 38.8 RetinaNet [33] ResNet-101 39.1 DetNet [31] DetNet-59 40.2 GA-Faster RCNN [56] ResNet-50 39.8 PANet [39] ResNet-50 39.8 TridentNet [29] ResNet-101 plement different backbones with the searched architecture from COCO in Table…”
PP-YOLO: An Effective and Efficient Implementation of Object Detector (2020, methodology): “Anchor-based methods are still the mainstream of object detection [33, 21, 31, 32, 1, 22, 28, 9, 45, 2, 5, 37, 20, 4, 15], which evolved from early proposal based detectors, such as Fast R-CNN [11].”
FoveaBox: Beyound Anchor-Based Object Detection (2019, methodology): “(c) At training phase, anchor-methods rely on the intersection-over-union (IoU) to deﬁne the positive/negative samples, which introduces additional computation and hyper-parameters for an object detection system [13].”
Probabilistic two-stage detection (2021, methodology): “Following Wang et al. (2019), we increase the positive IoU threshold in the second stage from 0 .”
SpatialFlow: Bridging All Tasks for Panoptic Segmentation (2019, background): “As demonstrated in the literature, the spatial information of objects can boost the performance of algorithms in object detection [12], [13], instance segmentation [14], [15], and semantic segmentation [16], [17].”
D2Det: Towards High Quality Object Detection and Instance Segmentation (2020, background): “In recent years, two-stage detection approaches [18, 43, 17, 44, 28, 36, 5, 46] have shown continuous performance improvements in terms of detection accuracy on standard benchmarks [33, 13].”