Filters








3,272 Hits in 4.3 sec

Adversarial Semantic Contour for Object Detection [article]

Yichi Zhang, Zijian Zhu, Xiao Yang, Jun Zhu
2021 arXiv   pre-print
To address this issue, we propose a novel method of Adversarial Semantic Contour (ASC) guided by object contour as prior.  ...  Modern object detectors are vulnerable to adversarial examples, which brings potential risks to numerous applications, e.g., self-driving car.  ...  Adversarial Patterns. Our methods (ASC) is based on the semantic boundaries on objects. For prior contour map mentioned in Sec.2.2, we acquire it with CDCL (Lin et al., 2020) .  ... 
arXiv:2109.15009v1 fatcat:prwu5fzszrahbg6sjo46ofaj6i

Knowledge-guided Semantic Computing Network [article]

Guangming Shi, Zhongqiang Zhang, Dahua Gao, Xuemei Xie, Yihao Feng, Xinrui Ma, Danhua Liu
2018 arXiv   pre-print
The object recognition process through the semantic tree only needs simple forward computing without training.  ...  The semantic tree is pre-defined to describe the spatial structural relations of different semantics, which just corresponds to the tree-like description of objects based on human knowledge.  ...  Shape Detection: In this step, we search for closed contours in the thinned image. The closed contours include circle or ellipse primitives.  ... 
arXiv:1810.00139v1 fatcat:flejnd2k6rcprfwdfo4xo4ygze

Semantic-shape Adaptive Feature Modulation for Semantic Image Synthesis [article]

Zhengyao Lv, Xiaoming Li, Zhenxing Niu, Bing Cao, Wangmeng Zuo
2022 arXiv   pre-print
Most previous methods focus on exploiting the given semantic map, which just captures an object-level layout for an image.  ...  Obviously, a fine-grained part-level semantic layout will benefit object details generation, and it can be roughly inferred from an object's shape.  ...  The part-level semantic layout is implied in the shape/contour of an object instance.  ... 
arXiv:2203.16898v1 fatcat:bvblwt3zdjebpnrqgz7uimeu2i

Adversarial Soft-detection-based Aggregation Network for Image Retrieval [article]

Jian Xu, Chunheng Wang, Cunzhao Shi, Baihua Xiao
2019 arXiv   pre-print
In this paper, we propose a novel adversarial soft-detection-based aggregation (ASDA) method free from bounding box annotations for image retrieval, based on adversarial detector and soft region proposal  ...  Our trainable adversarial detector generates semantic maps based on adversarial erasing strategy to preserve more discriminative and detailed information.  ...  The semantic map m 2 focuses on the body of buildings. The semantic map m K highlights the outer contour of Eiffel Tower.  ... 
arXiv:1811.07619v3 fatcat:cctz2l6ognbejg6xx3b555ebei

Can Giraffes Become Birds? An Evaluation of Image-to-image Translation for Data Generation [article]

Daniel V. Ruiz, Gabriel Salomon, Eduardo Todt
2020 arXiv   pre-print
For the quantitative analysis, a pre-trained Mask R-CNN was used for the detection and segmentation of birds on Pascal VOC, Caltech-UCSD Birds 200-2011, and our new dataset entitled FakeSet.  ...  In the present work, we investigate image-to-image translation using Generative Adversarial Networks (GANs) for generating new data, taking as a case study the morphing of giraffes images into bird images  ...  ACKNOWLEDGMENTS The authors would like to thank the Coordination for the Improvement of Higher Education Personnel (CAPES) for the Masters scholarship.  ... 
arXiv:2001.03637v2 fatcat:uirh5iwfsvdpvecs2t3nuqsc3q

SSMI: How to Make Objects of Interest Disappear without Accessing Object Detectors? [article]

Hui Xia, Rui Zhang, Zi Kang, Shuliang Jiang
2022 arXiv   pre-print
The maximum increase in new and disappearing labels is 16%, and the maximum decrease in mAP metrics for object detection is 36%.  ...  Most black-box adversarial attack schemes for object detectors mainly face two shortcomings: requiring access to the target model and generating inefficient adversarial examples (failing to make objects  ...  Liao et al. proposed the first adversarial attack against an unanchored object detection model and used high-level semantic information to generate effective adversarial examples [LWK + 20], but such methods  ... 
arXiv:2206.10809v1 fatcat:n3qfibeij5ac5cmqqwuv3wqkqy

Foreground-Aware Image Inpainting

Wei Xiong, Jiahui Yu, Zhe Lin, Jimei Yang, Xin Lu, Connelly Barnes, Jiebo Luo
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
These scenarios, however, are very important in practice, especially for applications such as the removal of distracting objects.  ...  We show that by such disentanglement, the contour completion model predicts reasonable contours of objects, and further substantially improves the performance of image inpainting.  ...  The refined contour is then fed to the contour discriminator for adversarial training.  ... 
doi:10.1109/cvpr.2019.00599 dblp:conf/cvpr/XiongYLYLBL19 fatcat:rvrnnlrygzeqzldezk7uihdz7e

VATLD: A Visual Analytics System to Assess, Understand and Improve Traffic Light Detection [article]

Liang Gou, Lincan Zou, Nanxiang Li, Michael Hofmann, Arvind Kumar Shekar, Axel Wendt, Liu Ren
2020 arXiv   pre-print
Traffic light detection is crucial for environment perception and decision-making in autonomous driving.  ...  The disentangled representation learning extracts data semantics to augment human cognition with human-friendly visual summarization, and the semantic adversarial learning efficiently exposes interpretable  ...  In this way, 1242 adversarial objects were obtained and mixed with global adversarial objects (3401) for adversarial training and validation.  ... 
arXiv:2009.12975v1 fatcat:35gtds7y6rcjxm7tvnk2orxgaa

Image-to-image Translation via Contour-consistency Networks

Hsiang-Ying Wang, Hsin-Chun Lin, Chih-Hsien Hsia, Natnuntnita Siriphockpirom, Hsien-I Lin, Yung-Yao Chen
2022 Sensors and materials  
In this paper, a novel framework for image-to-image translation, in which contourconsistency networks are used to solve the problem of inconsistency between the contours of generated and original images  ...  The objective of this study was to address the lack of an adequate training set.  ...  This dataset contains 24966 images, each of which has a pixel-level semantic annotation; the dataset is therefore popular for semantic segmentation.  ... 
doi:10.18494/sam3493 fatcat:mtt5fwi6cnbefescalswwkugoa

Modeling and Optimization of Semantic Segmentation for Track Bed Foreign Object Based on Attention Mechanism

Haoran Song, Shengchun Wang, Zichen Gu, Peng Dai, Xinyu Du, Yu Cheng
2021 IEEE Access  
To solve the problem, we propose an anomaly detection method for the ballastless track bed, which is based on semantic segmentation.  ...  Firstly, we put forward the RFODLab semantic segmentation network according to the randomness of foreign objects distribution, and a small proportion of target pixels in the track image.  ...  RFODLab SEMANTIC SEGMENTATION NETWORK For more effective anomaly detection, higher requirements are raised for the efficiency of segmenting and extracting foreign objects through the semantic segmentation  ... 
doi:10.1109/access.2021.3087705 fatcat:b7ctdtayhvb6lcvq7v2khjwmlu

Liver Semantic Segmentation Algorithm Based on Improved Deep Adversarial Networks in combination of Weighted Loss Function on Abdominal CT Images

Kaijian XIA, Hongsheng YIN, Pengjiang QIAN, Yizhang JIANG, Shuihua WANG
2019 IEEE Access  
segmentation efficiency while ensuring the space consistency of the semantics segmentation for abdominal CT images.  ...  INDEX TERMS Semantic segmentation, generation adversarial networks, weighted loss function, multi-scale features, game adversarial, atrous space pyramid pooling.  ...  Therefore, the establishment of a robust, objective, repeatable, efficient and high-accuracy method for detecting liver lesions has important clinical significance for the prevention and treatment of liver  ... 
doi:10.1109/access.2019.2929270 fatcat:2knqtanfnffnvfhz4ifpogl7xu

Improving Lesion Segmentation for Diabetic Retinopathy Using Adversarial Learning [chapter]

Qiqi Xiao, Jiaxu Zou, Muqiao Yang, Alex Gaudio, Kris Kitani, Asim Smailagic, Pedro Costa, Min Xu
2019 Lecture Notes in Computer Science  
We utilize the HEDNet edge detector to solve a semantic segmentation task on this dataset, and then propose an end-to-end system for pixel-level segmentation of DR lesions by incorporating HEDNet into  ...  DR lesions can be challenging to identify in fundus images, and automatic DR detection systems can offer strong clinical value.  ...  Unlike traditional edge detectors, HEDNet can generate semantically meaningful edge maps that identify object contours.  ... 
doi:10.1007/978-3-030-27272-2_29 fatcat:beoj2vn4ancppmgwdarur73hvq

Can Giraffes Become Birds? An Evaluation of Image-to-image Translation for Data Generation

Daniel Ruiz, Gabriel Salomon, Eduardo Todt
2020 Anais do Computer on the Beach  
The generateddataset achieved detection and segmentation results close tothe real datasets, suggesting that the generated images are realisticenough to be detected and segmented by a state-of-the-art deepneural  ...  ABSTRACTThere is an increasing interest in image-to-image translation withapplications ranging from generating maps from satellite images tocreating entire clothes' images from only contours.  ...  ACKNOWLEDGMENTS The authors would like to thank the Coordination for the Improvement of Higher Education Personnel (CAPES) for the Masters scholarship.  ... 
doi:10.14210/cotb.v11n1.p176-182 fatcat:2ul2kalmsfdklo6c7p7cqbjlxy

Dynamic Detection and Recognition of Objects Based on Sequential RGB Images

Shuai Dong, Zhihua Yang, Wensheng Li, Kun Zou
2021 Future Internet  
Many applications require fast, reliable, and dynamic detection and recognition for the objects on conveyors.  ...  MVRFFNet is a generalized zero-shot learning (GZSL) framework based on the Wasserstein generative adversarial network for 3D object recognition.  ...  [39] designed a semantic-preserving adversarial embedding network to avoid the loss of semantic information. Liu et al.  ... 
doi:10.3390/fi13070176 fatcat:wm6tolhqyfahjnp6quhubulu6y

Foreground-aware Image Inpainting [article]

Wei Xiong, Jiahui Yu, Zhe Lin, Jimei Yang, Xin Lu, Connelly Barnes and Jiebo Luo
2019 arXiv   pre-print
These scenarios, however, are very important in practice, especially for applications such as the removal of distracting objects.  ...  We show that by such disentanglement, the contour completion model predicts reasonable contours of objects, and further substantially improves the performance of image inpainting.  ...  The refined contour is then fed to the contour discriminator for adversarial training.  ... 
arXiv:1901.05945v3 fatcat:tn6kcl2ihbblli3nbaja7bc5ue
« Previous Showing results 1 — 15 out of 3,272 results