ASAN: Self-Attending and Semantic Activating Network towards Better Object Detection

Xinyu ZHU; Jun ZHANG; Gengsheng CHEN

doi:10.1587/transinf.2019EDP7164

ASAN: Self-Attending and Semantic Activating Network towards Better Object Detection

Xinyu ZHU, Jun ZHANG, Gengsheng CHEN

Full Text Views

0

Cite this

Summary :

Recent top-performing object detectors usually depend on a two-stage approach, which benefits from its region proposal and refining practice but suffers low detection speed. By contrast, one-stage approaches have the advantage of high efficiency while sacrifice their accuracies to some extent. In this paper, we propose a novel single-shot object detection network which inherits the merits of both. Motivated by the idea of semantic enrichment to the convolutional features within a typical deep detector, we propose two novel modules: 1) by modeling the semantic interactions between channels and the long-range dependencies between spatial positions, the self-attending module generates both channel and position attention, and enhance the original convolutional features in a self-guided manner; 2) leveraging the class-discriminative localization ability of classification-trained CNN, the semantic activating module learns a semantic meaningful convolutional response which augments low-level convolutional features with strong class-specific semantic information. The so called self-attending and semantic activating network (ASAN) achieves better accuracy than two-stage methods and is able to fulfil real-time processing. Comprehensive experiments on PASCAL VOC indicates that ASAN achieves state-of-the-art detection performance with high efficiency.

Publication: IEICE TRANSACTIONS on Information Vol.E103-D No.3 pp.648-659

Publication Date: 2020/03/01

Publicized: 2019/11/25

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2019EDP7164

Type of Manuscript: PAPER

Category: Image Recognition, Computer Vision

Authors

Xinyu ZHU
  Fudan University
Jun ZHANG
  Fudan University
Gengsheng CHEN
  Fudan University

Keyword

object detection, self-attending, semantic activating

Cite this

Copy

Xinyu ZHU, Jun ZHANG, Gengsheng CHEN, "ASAN: Self-Attending and Semantic Activating Network towards Better Object Detection" in IEICE TRANSACTIONS on Information, vol. E103-D, no. 3, pp. 648-659, March 2020, doi: 10.1587/transinf.2019EDP7164.
Abstract: Recent top-performing object detectors usually depend on a two-stage approach, which benefits from its region proposal and refining practice but suffers low detection speed. By contrast, one-stage approaches have the advantage of high efficiency while sacrifice their accuracies to some extent. In this paper, we propose a novel single-shot object detection network which inherits the merits of both. Motivated by the idea of semantic enrichment to the convolutional features within a typical deep detector, we propose two novel modules: 1) by modeling the semantic interactions between channels and the long-range dependencies between spatial positions, the self-attending module generates both channel and position attention, and enhance the original convolutional features in a self-guided manner; 2) leveraging the class-discriminative localization ability of classification-trained CNN, the semantic activating module learns a semantic meaningful convolutional response which augments low-level convolutional features with strong class-specific semantic information. The so called self-attending and semantic activating network (ASAN) achieves better accuracy than two-stage methods and is able to fulfil real-time processing. Comprehensive experiments on PASCAL VOC indicates that ASAN achieves state-of-the-art detection performance with high efficiency.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019EDP7164/_p

Copy

@ARTICLE{e103-d_3_648,
author={Xinyu ZHU, Jun ZHANG, Gengsheng CHEN, },
journal={IEICE TRANSACTIONS on Information},
title={ASAN: Self-Attending and Semantic Activating Network towards Better Object Detection},
year={2020},
volume={E103-D},
number={3},
pages={648-659},
abstract={Recent top-performing object detectors usually depend on a two-stage approach, which benefits from its region proposal and refining practice but suffers low detection speed. By contrast, one-stage approaches have the advantage of high efficiency while sacrifice their accuracies to some extent. In this paper, we propose a novel single-shot object detection network which inherits the merits of both. Motivated by the idea of semantic enrichment to the convolutional features within a typical deep detector, we propose two novel modules: 1) by modeling the semantic interactions between channels and the long-range dependencies between spatial positions, the self-attending module generates both channel and position attention, and enhance the original convolutional features in a self-guided manner; 2) leveraging the class-discriminative localization ability of classification-trained CNN, the semantic activating module learns a semantic meaningful convolutional response which augments low-level convolutional features with strong class-specific semantic information. The so called self-attending and semantic activating network (ASAN) achieves better accuracy than two-stage methods and is able to fulfil real-time processing. Comprehensive experiments on PASCAL VOC indicates that ASAN achieves state-of-the-art detection performance with high efficiency.},
keywords={},
doi={10.1587/transinf.2019EDP7164},
ISSN={1745-1361},
month={March},}

Copy

TY - JOUR
TI - ASAN: Self-Attending and Semantic Activating Network towards Better Object Detection
T2 - IEICE TRANSACTIONS on Information
SP - 648
EP - 659
AU - Xinyu ZHU
AU - Jun ZHANG
AU - Gengsheng CHEN
PY - 2020
DO - 10.1587/transinf.2019EDP7164
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2020
AB - Recent top-performing object detectors usually depend on a two-stage approach, which benefits from its region proposal and refining practice but suffers low detection speed. By contrast, one-stage approaches have the advantage of high efficiency while sacrifice their accuracies to some extent. In this paper, we propose a novel single-shot object detection network which inherits the merits of both. Motivated by the idea of semantic enrichment to the convolutional features within a typical deep detector, we propose two novel modules: 1) by modeling the semantic interactions between channels and the long-range dependencies between spatial positions, the self-attending module generates both channel and position attention, and enhance the original convolutional features in a self-guided manner; 2) leveraging the class-discriminative localization ability of classification-trained CNN, the semantic activating module learns a semantic meaningful convolutional response which augments low-level convolutional features with strong class-specific semantic information. The so called self-attending and semantic activating network (ASAN) achieves better accuracy than two-stage methods and is able to fulfil real-time processing. Comprehensive experiments on PASCAL VOC indicates that ASAN achieves state-of-the-art detection performance with high efficiency.
ER -