Region proposal mechanisms are essential for existing deep learning
approaches to object detection in images. Although they can generally achieve a
good detection performance under normal circumstances, their recall in a scene
with extreme cases is unacceptably low. This is mainly because bounding box
annotations contain much environment noise information, and non-maximum
suppression (NMS) is required to select target boxes. Therefore, in this paper,
we propose the first anchor-free and NMS-free object detection model called
weakly supervised multimodal annotation segmentation (WSMA-Seg), which utilizes
segmentation models to achieve an accurate and robust object detection without
NMS. In WSMA-Seg, multimodal annotations are proposed to achieve an
instance-aware segmentation using weakly supervised bounding boxes; we also
develop a run-data-based following algorithm to trace contours of objects. In
addition, we propose a multi-scale pooling segmentation (MSP-Seg) as the
underlying segmentation model of WSMA-Seg to achieve a more accurate
segmentation and to enhance the detection accuracy of WSMA-Seg. Experimental
results on multiple datasets show that the proposed WSMA-Seg approach
outperforms the state-of-the-art detectors.
%0 Generic
%1 cheng2019segmentation
%A Cheng, Zehua
%A Wu, Yuxiang
%A Xu, Zhenghua
%A Lukasiewicz, Thomas
%A Wang, Weiyang
%D 2019
%K arch backbone detection loss segmentation
%T Segmentation is All You Need
%U http://arxiv.org/abs/1904.13300
%X Region proposal mechanisms are essential for existing deep learning
approaches to object detection in images. Although they can generally achieve a
good detection performance under normal circumstances, their recall in a scene
with extreme cases is unacceptably low. This is mainly because bounding box
annotations contain much environment noise information, and non-maximum
suppression (NMS) is required to select target boxes. Therefore, in this paper,
we propose the first anchor-free and NMS-free object detection model called
weakly supervised multimodal annotation segmentation (WSMA-Seg), which utilizes
segmentation models to achieve an accurate and robust object detection without
NMS. In WSMA-Seg, multimodal annotations are proposed to achieve an
instance-aware segmentation using weakly supervised bounding boxes; we also
develop a run-data-based following algorithm to trace contours of objects. In
addition, we propose a multi-scale pooling segmentation (MSP-Seg) as the
underlying segmentation model of WSMA-Seg to achieve a more accurate
segmentation and to enhance the detection accuracy of WSMA-Seg. Experimental
results on multiple datasets show that the proposed WSMA-Seg approach
outperforms the state-of-the-art detectors.
@misc{cheng2019segmentation,
abstract = {Region proposal mechanisms are essential for existing deep learning
approaches to object detection in images. Although they can generally achieve a
good detection performance under normal circumstances, their recall in a scene
with extreme cases is unacceptably low. This is mainly because bounding box
annotations contain much environment noise information, and non-maximum
suppression (NMS) is required to select target boxes. Therefore, in this paper,
we propose the first anchor-free and NMS-free object detection model called
weakly supervised multimodal annotation segmentation (WSMA-Seg), which utilizes
segmentation models to achieve an accurate and robust object detection without
NMS. In WSMA-Seg, multimodal annotations are proposed to achieve an
instance-aware segmentation using weakly supervised bounding boxes; we also
develop a run-data-based following algorithm to trace contours of objects. In
addition, we propose a multi-scale pooling segmentation (MSP-Seg) as the
underlying segmentation model of WSMA-Seg to achieve a more accurate
segmentation and to enhance the detection accuracy of WSMA-Seg. Experimental
results on multiple datasets show that the proposed WSMA-Seg approach
outperforms the state-of-the-art detectors.},
added-at = {2019-06-23T22:38:19.000+0200},
author = {Cheng, Zehua and Wu, Yuxiang and Xu, Zhenghua and Lukasiewicz, Thomas and Wang, Weiyang},
biburl = {https://www.bibsonomy.org/bibtex/2463ee95f85666ec9a784be6ef366adcd/nmatsuk},
description = {Segmentation is All You Need},
interhash = {d0a13ce3b4a36e7abe2eafcfa413f049},
intrahash = {463ee95f85666ec9a784be6ef366adcd},
keywords = {arch backbone detection loss segmentation},
note = {cite arxiv:1904.13300Comment: 10 Pages},
timestamp = {2019-06-23T22:38:19.000+0200},
title = {Segmentation is All You Need},
url = {http://arxiv.org/abs/1904.13300},
year = 2019
}