CoupleNet: Coupling Global Structure with Local Parts for Object Detection

By Yousong Zhu, Chaoyang Zhao, Jinqiao Wang, Xu Zhao, Yi Wu, Hanqing Lu. 2017

271 citations recorded by the source corpus.

The region-based Convolutional Neural Network (CNN) detectors such as Faster R-CNN or R-FCN have already shown promising results for object detection by combining the region proposal subnetwork and the classification subnetwork together. Although R-FCN has achieved higher detection speed while keeping the detection performance, the global structure information is ignored by the position-sensitive score maps. To fully explore the local and global properties, in this paper, we propose a novel fully convolutional network, named as CoupleNet, to couple the global structure with local parts for object detection. Specifically, the object proposals obtained by the Region Proposal Network (RPN) are fed into the the coupling module which consists of two branches. One branch adopts the position-sensitive RoI (PSRoI) pooling to capture the local part information of the object, while the other employs the RoI pooling to encode the global and context information. Next, we design different coupling strategies and normalization ways to make full use of the complementary advantages between the global and local branches. Extensive experiments demonstrate the effectiveness of our approach. We achieve state-of-the-art results on all three challenging datasets, i.e. a mAP of 82.7% on VOC07, 80.4% on VOC12, and 34.4% on COCO. Codes will be made publicly available1.

Subfields: 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 Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Ling Shao
3. Xiaogang Wang
4. Dacheng Tao
5. Ming-Hsuan Yang
6. Yi Yang
7. Qi Tian
8. Chunhua Shen

Citations in the corpus

• Bridging the Gap Between Anchor-Based and Anchor-Free Detection via Adaptive Training Sample Selection (2019, neutral): “0 CoupleNet [74] trainval ResNet-101 34.”
• BAM: Bottleneck Attention Module (2018, background): “This phenomenon can be interpreted in terms of contextual reasoning, which is widely exploited in dense prediction tasks [4, 5, 34, 46, 49].”
• StairNet: Top-Down Semantic Aggregation for Accurate One Shot Detection (2017, background): “CoupleNet [42] introduces global FCN branch to extract global semantics.”
• CornerNet: Detecting Objects as Paired Keypoints (2018, neutral): “3 CoupleNet (Zhu et al., 2017) ResNet-101 34.”
• Deep High-Resolution Representation Learning for Visual Recognition (2019, neutral): “CoupleNet [194] ResNet-101 33.”
• CenterNet: Keypoint Triplets for Object Detection (2019, background): “, a [52, 22] focus on the architecture design, a [1, 10, 37, 47] focus on the contextual relationship, and a [23, 3] focus on the multi-scale unification.”
• Bottom-Up Object Detection by Grouping Extreme and Center Points (2019, background): “The detection-by-classification idea is intuitive and keeps the best performance so far [6,7, 19,20,24,27,37,45,46,54].”
• Feature Selective Anchor-Free Module for Single-Shot Object Detection (2019, methodology): “Method Backbone AP AP50 AP75 APS APM APL Multi-shot detectors CoupleNet [42]”
• Deep Feature Pyramid Reconfiguration for Object Detection (2018, neutral): “9 CoupleNet[49] trainval35k ∼ 1000× 600 ResNet-101 54.”
• ScratchDet: Exploring to Train Single-Shot Object Detectors from Scratch (2018, unknown): “6 CoupleNet[42] ResNet-101 ∼ 1000× 600 8.”
• Single-Shot Bidirectional Pyramid Networks for High-Quality Object Detection (2018, unknown): “7 CoupleNet [24] ResNet-101 ∼ 1000 × 600 8.”
• Enriched Feature Guided Refinement Network for Object Detection (2019, methodology): “CoupleNet [45] ResNet101 1000× 600 121ms 34.”
• Sampling Techniques for Large-Scale Object Detection From Sparsely Annotated Objects (2018, methodology): “ance of a single model with single scale testing on the validation split with bells and whistles. validation mAP Baseline 64.5 + Part-aware sampling 65.2 (+0.7) + 16 epochs 65.8 (+0.6) + Context head [28] 66.0 (+0.2) + SENet-154 and additional anchors 67.5 (+1.5) Table 9: Ensemble of models with tes”
• Squeeze and Excitation Rank Faster R-CNN for Ship Detection in SAR Images (2019, methodology): “Second, using modified RPN architectures to improve the first-stage performance and generate more accurate region proposals [16], [17] for the second-stage classification.”
• Recent Advances in Deep Learning for Object Detection (2019, methodology): “6 – CoupleNet [125] ResNet-101 2017 ∼ 600 × 1000 82.”
• LinkNet: Relational Embedding for Scene Graph (2018, unknown): “It is well known that fusing global and local information plays an important role in numerous visual tasks [8, 23, 39, 6, 40, 36].”
• Receptive Field Block Net for Accurate and Fast Object Detection (2017, methodology): “CoupleNet [43] adopts local information partly based on R-FCN [19].”