Filters








447 Hits in 5.4 sec

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  ...  We present ALADIN (All Layer AdaIN); a novel architecture for searching images based on the similarity of their artistic style.  ...  Finally, we demonstrate the benefit of contrastive learning for weakly supervised learning, by baselining performance against softmax loss for an n−way classification using project group membership as  ... 
arXiv:2103.09776v1 fatcat:gddjgr4zcnetlp26aihmy2aerq

Front Matter: Volume 12032

Ivana Išgum, Olivier Colliot
2022 Medical Imaging 2022: Image Processing  
segmentation of anatomical structures in fetal torso ultrasound [12032-114] 3A Calcium scoring in low-dose ungated chest CT scans using convolutional long-short term memory networks [12032-115] 3B Weakly  ...  diffusion-weighted data using tagged magnetic resonance imaging [ models for organ contouring in head and neck radiotherapy [12032-13] 0G Automatic classification of MRI contrasts using a deep siamese  ... 
doi:10.1117/12.2638192 fatcat:ikfgnjefaba2tpiamxoftyi6sa

Specialized Decision Surface and Disentangled Feature for Weakly-Supervised Polyphonic Sound Event Detection [article]

Liwei Lin, Xiangdong Wang, Hong Liu, Yueliang Qian
2020 arXiv   pre-print
In this paper, a special decision surface for the weakly-supervised sound event detection (SED) and a disentangled feature (DF) for the multi-label problem in polyphonic SED are proposed.  ...  We present a method of generating instance-level probabilities for the embedding level approaches which tend to perform better than the instance-level approaches in terms of bag-level classification but  ...  SED: a fully convolutional network with a GMP module [23] , a joint detection-classification (JDC) model with a ATP module [24] , a convolutional neural network (CNN) based model with a GAP module  ... 
arXiv:1905.10091v6 fatcat:rpm2m4tskjdirnq6vfw2pzqltm

Group-disentangled Representation Learning with Weakly-Supervised Regularization [article]

Linh Tran, Amir Hosein Khasahmadi, Aditya Sanghi, Saeid Asgari
2021 arXiv   pre-print
We investigate learning group-disentangled representations for groups of factors with weak supervision.  ...  Further, we demonstrate that learning group-disentangled representations improve upon downstream tasks, including fair classification and 3D shape-related tasks such as reconstruction, classification,  ...  ACKNOWLEDGMENTS We thank Hooman Shayani and Tonya Custis for useful discussions and comments on the paper.  ... 
arXiv:2110.12185v1 fatcat:ywcwv65qbrbk7b57mrzj27x2ui

2021 Index IEEE Transactions on Multimedia Vol. 23

2021 IEEE transactions on multimedia  
The Author Index contains the primary entry for each item, listed under the first author's name.  ...  ., +, TMM 2021 3306-3317 Anisotropic Graph Convolutional Network for Semi-Supervised Learning.  ...  ., +, TMM 2021 4014-4026 Anisotropic magnetoresistance Anisotropic Graph Convolutional Network for Semi-Supervised Learning.  ... 
doi:10.1109/tmm.2022.3141947 fatcat:lil2nf3vd5ehbfgtslulu7y3lq

Weakly Supervised Visual Semantic Parsing [article]

Alireza Zareian, Svebor Karaman, Shih-Fu Chang
2020 arXiv   pre-print
Additionally, we propose the first graph-based weakly supervised learning framework, based on a novel graph alignment algorithm, which enables training without bounding box annotations.  ...  Through extensive experiments, we show that VSPNet outperforms weakly supervised baselines significantly and approaches fully supervised performance, while being several times faster.  ...  Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation here on.  ... 
arXiv:2001.02359v2 fatcat:mdxk3kd77bgploitjm3j7obtqa

Explanatory Graphs for CNNs [article]

Quanshi Zhang, Xin Wang, Ruiming Cao, Ying Nian Wu, Feng Shi, Song-Chun Zhu
2018 arXiv   pre-print
Each filter in a conv-layer of a CNN for object classification usually represents a mixture of object parts.  ...  We develop a simple yet effective method to disentangle object-part pattern components from each filter.  ...  INTRODUCTION Convolutional neural networks (CNNs) [9] , [12] , [15] have exhibited superior performance in various visual tasks, for example, object classification and detection.  ... 
arXiv:1812.07997v1 fatcat:vq7pzmcqgbhzvlulnt4ypgt5mq

Weakly Supervised Visual Semantic Parsing

Alireza Zareian, Svebor Karaman, Shih-Fu Chang
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Additionally, we propose the first graphbased weakly supervised learning framework, based on a novel graph alignment algorithm, which enables training without bounding box annotations.  ...  Through extensive experiments, we show that VSPNET outperforms weakly supervised baselines significantly and approaches fully supervised performance, while being several times faster.  ...  Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation here on.  ... 
doi:10.1109/cvpr42600.2020.00379 dblp:conf/cvpr/ZareianKC20 fatcat:h3v5phfq3ze4refvwhfntufbri

Interpreting CNN Knowledge via an Explanatory Graph [article]

Quanshi Zhang, Ruiming Cao, Feng Shi, Ying Nian Wu, Song-Chun Zhu
2017 arXiv   pre-print
More importantly, we learn the explanatory graph for a pre-trained CNN in an unsupervised manner, i.e., without a need of annotating object parts.  ...  Considering that each filter in a conv-layer of a pre-trained CNN usually represents a mixture of object parts, we propose a simple yet efficient method to automatically disentangles different part patterns  ...  These methods did not use part annotations, and only used object boxes for learning/selection.  ... 
arXiv:1708.01785v3 fatcat:vct2ttai75bplhqeghapwwwyli

WildGait: Learning Gait Representations from Raw Surveillance Streams [article]

Adrian Cosma, Emilian Radoi
2021 arXiv   pre-print
As such, we propose a novel weakly supervised learning framework, WildGait, which consists of training a Spatio-Temporal Graph Convolutional Network on a large number of automatically annotated skeleton  ...  Existing methods for gait recognition require cooperative gait scenarios, in which a single person is walking multiple times in a straight line in front of a camera.  ...  Similar to us, applies a graph convolutional network to process skeleton sequences, but use a final pyramid pooling layer for recognition.  ... 
arXiv:2105.05528v5 fatcat:vab4fmeh2rdcri5pttq723njwu

2021 Index IEEE Transactions on Image Processing Vol. 30

2021 IEEE Transactions on Image Processing  
The Author Index contains the primary entry for each item, listed under the first author's name.  ...  ., +, TIP 2021 6648-6658 AP-CNN: Weakly Supervised Attention Pyramid Convolutional Neural Network for Fine-Grained Visual Classification.  ...  ., +, TIP 2021 2810-2825 AP-CNN: Weakly Supervised Attention Pyramid Convolutional Neural Network for Fine-Grained Visual Classification.  ... 
doi:10.1109/tip.2022.3142569 fatcat:z26yhwuecbgrnb2czhwjlf73qu

Aligned to the Object, not to the Image: A Unified Pose-aligned Representation for Fine-grained Recognition [article]

Pei Guo, Ryan Farrell
2018 arXiv   pre-print
network.  ...  The effectiveness of this paradigm relative to competing methods suggests the critical importance of disentangling pose and appearance for continued progress in fine-grained recognition.  ...  We contrast the pose-aligned regions with weakly-supervised regions that are generated in a purely data-driven fashion and with "axis-aligned" rectangular bounding boxes centered around a keypoint or landmark  ... 
arXiv:1801.09057v4 fatcat:nwuvizkmebfbrhonnkfwkuhxdy

DL4SciVis: A State-of-the-Art Survey on Deep Learning for Scientific Visualization [article]

Chaoli Wang, Jun Han
2022 arXiv   pre-print
In this paper, we survey related deep learning (DL) works in SciVis, specifically in the direction of DL4SciVis: designing DL solutions for solving SciVis problems.  ...  We classify and discuss these works along six dimensions: domain setting, research task, learning type, network architecture, loss function, and evaluation metric.  ...  The authors would like to thank the anonymous reviewers for their insightful comments.  ... 
arXiv:2204.06504v1 fatcat:33fc2smtuffwll6pghdffbebi4

VAE-CE: Visual Contrastive Explanation using Disentangled VAEs [article]

Yoeri Poels, Vlado Menkovski
2021 arXiv   pre-print
We build the model using a disentangled VAE, extended with a new supervised method for disentangling individual dimensions.  ...  Such concepts can also be highly useful for interpreting the model's classifications.  ...  we can expect well-disentangled representations, of explanation in the context of image classification using showing that strong inductive biases or supervision are a neural networks, and disentanglement  ... 
arXiv:2108.09159v1 fatcat:onejltviqzavfm6phethxxux3q

MakeupBag: Disentangling Makeup Extraction and Application [article]

Dokhyam Hoshen
2020 arXiv   pre-print
We solve makeup disentanglement and facial makeup application as separable objectives, in contrast to other current deep methods that entangle the two tasks.  ...  This paper introduces MakeupBag, a novel method for automatic makeup style transfer.  ...  Weakly supervised segmentation MakeupBag addresses makeup extraction as a weaklysupervised segmentation task.  ... 
arXiv:2012.02157v1 fatcat:icgnqmikgnaw5lrluqnmd6cygy
« Previous Showing results 1 — 15 out of 447 results