Filters








2,166 Hits in 5.4 sec

Coarse-to-fine Semantic Segmentation from Image-level Labels [article]

Longlong Jing, Yucheng Chen, Yingli Tian
2018 arXiv   pre-print
For each image, an initial coarse mask is first generated by a convolutional neural network-based unsupervised foreground segmentation model and then is enhanced by a graph model.  ...  In this paper, we propose a novel recursive coarse-to-fine semantic segmentation framework based on only image-level category labels.  ...  the coarse masks to fine, while the coarse masks are generated by an unsupervised foreground segmentation method and enhanced by a graph model.  ... 
arXiv:1812.10885v1 fatcat:uc7dsdf6jrfkfn32abfgxcwcpe

Detecting 11K Classes: Large Scale Object Detection Without Fine-Grained Bounding Boxes

Hao Yang, Hao Wu, Hao Chen
2019 2019 IEEE/CVF International Conference on Computer Vision (ICCV)  
We achieve this by utilizing the correlations between coarse-grained and fine-grained classes with shared backbone, soft-attention based proposal reranking, and a dual-level memory module.  ...  In this paper, we propose a semi-supervised large scale fine-grained detection method, which only needs bounding box annotations of a smaller number of coarsegrained classes and image-level labels of large  ...  For fully-supervised detection stream, we do not have proposal-level coarse-grained labels for the fine-grained classification data, and for weakly-supervised stream, we do not have image-level fine-grained  ... 
doi:10.1109/iccv.2019.00990 dblp:conf/iccv/YangWC19 fatcat:euiavla6uvfs3id5sobuul6v5y

Coarse-to-Fine Pre-training for Named Entity Recognition [article]

Mengge Xue, Bowen Yu, Zhenyu Zhang, Tingwen Liu, Yue Zhang, Bin Wang
2020 arXiv   pre-print
Then we leverage thegazetteer-based distant supervision strategy totrain the model extract coarse-grained typedentities.  ...  To this end, we proposea NER-specific pre-training framework to in-ject coarse-to-fine automatically mined entityknowledge into pre-trained models.  ...  In this paper, we propose a Coarse-to-Fine Entity knowledge Enhanced (CoFEE) pre-training framework for NER task, aiming to gather and utilize knowledge related to named entities.  ... 
arXiv:2010.08210v1 fatcat:fdrrnwjsnffk7mep46muushu7y

Coarse-to-Fine Pseudo-Labeling Guided Meta-Learning for Inexactly-Supervised Few-Shot Classification [article]

Jinhai Yang, Hua Yang, Lin Chen
2020 arXiv   pre-print
Accordingly, we propose a Coarse-to-Fine (C2F) pseudo-labeling process to construct pseudo-tasks from coarsely-labeled data by grouping each coarse-class into pseudo-fine-classes via similarity matching  ...  In this paper, we present a new problem named inexactly-supervised meta-learning to alleviate such limitation, focusing on tackling few-shot classification tasks with only coarse-grained supervision.  ...  Coarse-to-Fine Pseudo-Labeling Unlike the pseudo-labeling in the usual sense [25] , coarse-to-fine pseudo-labeling takes as input a coarsely-annotated image and assigns a fine-grained pseudo-label to  ... 
arXiv:2007.05675v2 fatcat:czev6hm2mfep7dascne5fioeza

ALADIN: All Layer Adaptive Instance Normalization for Fine-grained Style Similarity [article]

Dan Ruta, Saeid Motiian, Baldo Faieta, Zhe Lin, Hailin Jin, Alex Filipkowski, Andrew Gilbert, John Collomosse
2021 arXiv   pre-print
ALADIN takes a weakly supervised approach to learning a representation for fine-grained style similarity of digital artworks, leveraging BAM-FG, a novel large-scale dataset of user generated content groupings  ...  ALADIN sets a new state of the art accuracy for style-based visual search over both coarse labelled style data (BAM) and BAM-FG; a new 2.62 million image dataset of 310,000 fine-grained style groupings  ...  Weakly supervised learning of fine-grained style. We present the first study into representation learning for fine-grained artistic style similarity, taking a weakly supervised approach.  ... 
arXiv:2103.09776v1 fatcat:gddjgr4zcnetlp26aihmy2aerq

Weakly-Supervised Semantic Segmentation by Iteratively Mining Common Object Features

Xiang Wang, Shaodi You, Xi Li, Huimin Ma
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
Weakly-supervised semantic segmentation under image tags supervision is a challenging task as it directly associates high-level semantic to low-level appearance.  ...  Experimental results on Pascal VOC 2012 dataset demonstrate that the proposed method outperforms previous state-of-the-art methods by a large margin.  ...  Introduction Weakly-supervised semantic segmentation under image tags supervision is to perform a pixel-wise segmentation of an image, providing only the labels of existing semantic objects in the image  ... 
doi:10.1109/cvpr.2018.00147 dblp:conf/cvpr/WangYLM18 fatcat:e2hkz3zp4rabzkvs5pcal6xrai

Semi-supervised fine-grained image categorization using transfer learning with hierarchical multi-scale adversarial networks

Peng Chen, Peng Li, Qing Lia, Dezheng Zhang
2019 IEEE Access  
With this approach, cross domain features are extracted by advanced deep encoders coarsely.  ...  Fine-grained image categorization is still a challenging computer vision problem in recent years.  ...  Due to the utilization of the knowledge and ''coarse'' labels in source domain, AMAN can obtain a higher accuracy with very fewer training data (only 20%) than existing weakly-supervised methods (e.g.  ... 
doi:10.1109/access.2019.2934476 fatcat:stu53kvrs5gc3ortvtmijywpda

Weakly-Supervised Semantic Segmentation by Iteratively Mining Common Object Features [article]

Xiang Wang, Shaodi You, Xi Li, Huimin Ma
2018 arXiv   pre-print
Weakly-supervised semantic segmentation under image tags supervision is a challenging task as it directly associates high-level semantic to low-level appearance.  ...  Experimental results on Pascal VOC 2012 dataset demonstrate that the proposed method outperforms previous state-of-the-art methods by a large margin.  ...  Introduction Weakly-supervised semantic segmentation under image tags supervision is to perform a pixel-wise segmentation of an image, providing only the labels of existing semantic objects in the image  ... 
arXiv:1806.04659v1 fatcat:qjtigjd34rax7m7flropvmp4gy

Exploiting Category Similarity-based Distributed Labeling for Fine-Grained Visual Classification

Pengzhen Du, Zeren Sun, Yazhou Yao, Zhenmin Tang
2020 IEEE Access  
The second group is weakly supervised approaches. Different from strongly supervised methods, weakly supervised methods cease to use bounding boxes and part annotations.  ...  Instead, methods in this group only require image-level labels during training [20] - [32] , [39] , [40] . Some of them also follow a part-based pipeline but in a weakly supervised manner.  ... 
doi:10.1109/access.2020.3030249 fatcat:k32y6jwqhfextmngzi5b5vbecm

CAMEL: A Weakly Supervised Learning Framework for Histopathology Image Segmentation [article]

Gang Xu, Zhigang Song, Zhuo Sun, Calvin Ku, Zhe Yang, Cancheng Liu, Shuhao Wang, Jianpeng Ma, Wei Xu
2019 arXiv   pre-print
In this research, we propose CAMEL, a weakly supervised learning framework for histopathology image segmentation using only image-level labels.  ...  Moreover, the generality of the automatic labeling methodology may benefit future weakly supervised learning studies for histopathology image analysis.  ...  This research is supported by National Natural Science Foundation of China (NSFC) (No.  ... 
arXiv:1908.10555v1 fatcat:sphx4sfvsrhblglauzpotenace

LSTD: A Low-Shot Transfer Detector for Object Detection [article]

Hao Chen, Yali Wang, Guoyou Wang, Yu Qiao
2018 arXiv   pre-print
knowledge respectively from source and target domains, in order to further enhance fine-tuning with a few target images.  ...  Recent advances in object detection are mainly driven by deep learning with large-scale detection benchmarks.  ...  On the contrary, both weakly-supervised and semi-supervised approaches require the full training set (i.e., weakly-supervised: all training images with only imagelevel labels, semi-supervised: a few fully-annotated  ... 
arXiv:1803.01529v1 fatcat:2dw66v2kvvbrjdy4lldnuh233u

Weakly Supervised Segmentation for Real-time Surgical Tool Tracking

Eung Joo Lee, William Plishker, Xinyang Liu, S Bhattacharyya, Raj Shekhar
2019 Healthcare technology letters  
To tackle this issue, the authors propose a weakly supervised method for surgical tool segmentation and tracking based on hybrid sensor systems.  ...  Electromagnetic (EM) tracking can be utilised for tool tracking, but the accuracy is often limited by magnetic interference.  ...  Semantic labels generated in the Semantic Labelling subsystem are used to train a DCNN for the Weakly Supervised Segmentation subsystem Fig. 2 2 Coarse seed generation using EM tracking a Hybrid sensor  ... 
doi:10.1049/htl.2019.0083 pmid:32038863 pmcid:PMC6952260 fatcat:7ekhd36cofd7dahcirtjesq4pm

A Survey of Visual Sensory Anomaly Detection [article]

Xi Jiang, Guoyang Xie, Jinbao Wang, Yong Liu, Chengjie Wang, Feng Zheng, Yaochu Jin
2022 arXiv   pre-print
Furthermore, we classify each kind of anomaly according to the level of supervision. Finally, we summarize the challenges and provide open directions for this community.  ...  In this survey, we are the first one to provide a comprehensive review of visual sensory AD and category into three levels according to the form of anomalies.  ...  weakly supervised learning, i.e., just coarsely annotate the abnormal event in the video sequence and their task is to finely localize the anomaly event in each video frame. propose a dual-path based mutually-guided  ... 
arXiv:2202.07006v1 fatcat:2bqzmmrnjzggti5tcewa3mh3sa

A Mutual Bootstrapping Model for Automated Skin Lesion Segmentation and Classification [article]

Yutong Xie, Jianpeng Zhang, Yong Xia, Chunhua Shen
2020 arXiv   pre-print
This model consists of a coarse segmentation network (coarse-SN), a mask-guided classification network (mask-CN), and an enhanced segmentation network (enhanced-SN).  ...  On one hand, the coarse-SN generates coarse lesion masks that provide a prior bootstrapping for mask-CN to help it locate and classify skin lesions accurately.  ...  [30] proposed a decoupled network for weakly-supervised segmentation, where the classspecific localization cues are transferred from the classifi- cation network to the segmentation network.  ... 
arXiv:1903.03313v4 fatcat:cv3ldlhts5gndpxb6ttmrlc3ya

Facial Component-Landmark Detection With Weakly-Supervised LR-CNN

Ruiheng Zhang, Chengpo Mu, Min Xu, Lixin Xu, Xiaofeng Xu
2019 IEEE Access  
Then, through weakly supervised learning, our LR-CNN model can be trained effectively with a small amount of finely labeled data and a large amount of generated weakly labeled data.  ...  We can handle the task with a small amount of finely labeled data.  ...  Coarse-to-fine shape searching (CFSS) [52] .  ... 
doi:10.1109/access.2018.2890573 fatcat:czjdhb4xujeq7hf3q6wzwdnqiu
« Previous Showing results 1 — 15 out of 2,166 results