167 Hits in 4.3 sec

AnchorNet: A Weakly Supervised Network to Learn Geometry-sensitive Features For Semantic Matching [article]

David Novotny, Diane Larlus, Andrea Vedaldi
2017 arXiv   pre-print
In this work, we propose a deep network, termed AnchorNet, that produces image representations that are well-suited for semantic matching.  ...  weak image-level labels, the final representation successfully captures information about the object structure and improves results of state-of-the-art semantic matching methods such as the deformable spatial  ...  We would like to thank Xerox Research Center Europe and ERC 677195-IDIU for supporting this research.  ... 
arXiv:1704.04749v1 fatcat:kn7jl4dixbe2jgk5xm7cjlaiga

AnchorNet: A Weakly Supervised Network to Learn Geometry-Sensitive Features for Semantic Matching

David Novotny, Diane Larlus, Andrea Vedaldi
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In this work, we propose a deep network, termed AnchorNet, that produces image representations that are well-suited for semantic matching.  ...  weak image-level labels, the final representation successfully captures information about the object structure and improves results of state-of-the-art semantic matching methods such as the deformable spatial  ...  Finally, WarpNet [29] learns correspondences by exploiting the relationships within a fine-grained dataset.  ... 
doi:10.1109/cvpr.2017.306 dblp:conf/cvpr/NovotnyLV17 fatcat:4xivgkrkyjeqnpwxt3j3rthnc4

Locally Supervised Deep Hybrid Model for Scene Recognition

Sheng Guo, Weilin Huang, Limin Wang, Yu Qiao
2017 IEEE Transactions on Image Processing  
Firstly, we notice that the convolutional features capture local objects and fine structures of scene images, which yield important cues for discriminating ambiguous scenes, whereas these features are  ...  In this paper, we propose a novel Locally-Supervised Deep Hybrid Model (LS-DHM) that effectively enhances and explores the convolutional features for scene recognition.  ...  For the scene images, such local semantics are of importance for fine-grained categorization, but are required to increase their robustness by discarding explicit spatial information.  ... 
doi:10.1109/tip.2016.2629443 pmid:28113936 fatcat:m3p32azwd5bsvky3simmy5cqfi

SSNet: Learning Mid-Level Image Representation Using Salient Superpixel Network

Zhihang Ji, Fan Wang, Xiang Gao, Lijuan Xu, Xiaopeng Hu
2019 Applied Sciences  
In the standard bag-of-visual-words (BoVW) model, the burstiness problem of features and the ignorance of high-order information often weakens the discriminative power of image representation.  ...  For reducing the impact of burstiness occurred in the background region, we use the salient regions instead of the whole image to extract local features, and a fast saliency detection algorithm based on  ...  In [15] , Sharma et al. extend the notion of discriminative visual saliency by including discriminative spatial information and obtain a more discriminative image representation for visual classification  ... 
doi:10.3390/app10010140 fatcat:f45khoh22zdn7f7gz7wgfdo4w4

Deep Learning for Scene Classification: A Survey [article]

Delu Zeng, Minyu Liao, Mohammad Tavakolian, Yulan Guo, Bolei Zhou, Dewen Hu, Matti Pietikäinen, Li Liu
2021 arXiv   pre-print
This work was supported in part by grants from National Science  ...  feature refinement so that the network focuses on the most discriminative parts.  ...  Fine-tuning RGB-CNNs for depth images.  ... 
arXiv:2101.10531v2 fatcat:hwqw5so46ngxdlnfw7zynmpu6m

Video genre categorization and representation using audio-visual information

Bogdan Ionescu
2012 Journal of Electronic Imaging (JEI)  
We propose an audio-visual approach to video genre classification using content descriptors that exploit audio, color, temporal, and contour information.  ...  Also, experimental comparison as part of 1 the MediaEval 2011 benchmarking campaign demonstrated the superiority of the proposed audiovisual descriptors over other existing approaches.  ...  providing them with access to their animated movie database and for their support.  ... 
doi:10.1117/1.jei.21.2.023017 fatcat:ftmpjzlx5rdndbcmwiqx7jktga

Mining Personal Data Using Smartphones and Wearable Devices: A Survey

Muhammad Rehman, Chee Liew, Teh Wah, Junaid Shuja, Babak Daghighi
2015 Sensors  
In this study, we presented the details of data sources, design considerations, different application models, and data mining algorithms for PerDM in RCEs.  ...  Moreover, we explored the latest relevant literature to find the empirical evidences for evaluation criteria used for data mining algorithms in RCEs.  ...  In addition the authors would like to acknowledge Bright Spark Unit of University of Malaya for providing incentive support under grant no. BSP/APP/1634/2013.  ... 
doi:10.3390/s150204430 pmid:25688592 pmcid:PMC4367420 fatcat:3lcjmkqjeracbd6zxrzac4xcui

A Survey on Deep Visual Place Recognition

Carlo Masone, Barbara Caputo
2021 IEEE Access  
To overcome this problem they propose a fine-grained segmentation network with a high number of classes (≈ 10 2 − 10 3 ) that are learned in a self-supervised way by clustering the features extracted.  ...  SELECTING WHERE TO LOOK The problem of coping with visual clutter and distractors has inspired different solutions for guiding the visual inspection pipeline to focus on the most informative parts of the  ... 
doi:10.1109/access.2021.3054937 fatcat:hc5fp2z4g5fldl7imkt7ixq4z4

User Response Prediction in Online Advertising [article]

Zhabiz Gharibshah, Xingquan Zhu
2021 arXiv   pre-print
What type of data are available for user response prediction? How to predict user response in a reliable and/or transparent way?  ...  We propose a taxonomy to categorize state-of-the-art user response prediction methods, primarily focus on the current progress of machine learning methods used in different online platforms.  ...  In [53] authors extended the feature interaction modeling in factorization machines using a bilinear interaction method which combines inner product and Hadamard product together to generate a fine-grained  ... 
arXiv:2101.02342v2 fatcat:clgefamcd5fmbeg5ephizy3zqu

Show, Match and Segment: Joint Weakly Supervised Learning of Semantic Matching and Object Co-segmentation [article]

Yun-Chun Chen, Yen-Yu Lin, Ming-Hsuan Yang, Jia-Bin Huang
2020 arXiv   pre-print
Our model is end-to-end trainable and does not require supervision from manually annotated correspondences and object masks.  ...  We present an approach for jointly matching and segmenting object instances of the same category within a collection of images.  ...  discriminative parts (for object co-segmentation).  ... 
arXiv:1906.05857v2 fatcat:pmw2bgz7nzhl5hyjhxogp3nply

Privacy Intelligence: A Survey on Image Privacy in Online Social Networks [article]

Chi Liu, Tianqing Zhu, Jun Zhang, Wanlei Zhou
2021 arXiv   pre-print
Image sharing on online social networks (OSNs) has become an indispensable part of daily social activities, but it has also led to an increased risk of privacy invasion.  ...  A more intelligent environment for privacy-friendly OSN image sharing is in demand.  ...  The other goal is to perform more personalized and fine-grained encryption.  ... 
arXiv:2008.12199v2 fatcat:vxv6rsnyavesjiw2bbnc4jzsiy

Efficient Pairwise Neuroimage Analysis using the Soft Jaccard Index and 3D Keypoint Sets [article]

Laurent Chauvin, Kuldeep Kumar, Christian Desrosiers, William Wells III, Matthew Toews
2021 arXiv   pre-print
Software is provided for efficient fine-grained curation of large, generic image datasets.  ...  Our measure generalizes the Jaccard index to account for soft set equivalence (SSE) between keypoint elements, via an adaptive kernel framework modeling uncertainty in keypoint appearance and geometry.  ...  A memory-based model using efficient and robust algorithms for 3D keypoint extraction and indexing [11] , [66] allows for fine-grained comparisons between O(N 2 ) image pairs in O(N log N ) computational  ... 
arXiv:2103.06966v3 fatcat:2l4r56zz65e6nduzzgrav4mksa

Computational models of music perception and cognition II: Domain-specific music processing

Hendrik Purwins, Maarten Grachten, Perfecto Herrera, Amaury Hazan, Ricard Marxer, Xavier Serra
2008 Physics of Life Reviews  
Acknowledgments Special thanks to Hugo Solis for contributing to the sections on melody in this article. We are grateful to Fabien Gouyon for useful comments.  ...  We would like to thank Graham Coleman for proof reading the manuscript. This work is funded by EU Open FET IST-FP6-013123 (EmCAP) and the Spanish TIC project ProSeMus (TIN2006-14932-C02-01).  ...  Palmer and Krumhansl [1990] also show that listeners appear to have abstract knowledge of the accent structure of different kinds of meters, which become more fine-grained as a result of musical training  ... 
doi:10.1016/j.plrev.2008.03.005 fatcat:ntr3a2smfjd3dfrfz3musk5dhm

Image Matching from Handcrafted to Deep Features: A Survey

Jiayi Ma, Xingyu Jiang, Aoxiang Fan, Junjun Jiang, Junchi Yan
2020 International Journal of Computer Vision  
As a fundamental and critical task in various visual applications, image matching can identify then correspond the same or similar structure/content from two or more images.  ...  This encourages us to conduct a comprehensive and systematic review and analysis for those classical and latest techniques.  ...  Wang et al. (2014) proposed a novel deep ranking model to learn fine-grained image similarity.  ... 
doi:10.1007/s11263-020-01359-2 fatcat:a2epfaolwjfm5mcrsmn7g6sd7m

Reverse Engineering of Generative Models: Inferring Model Hyperparameters from Generated Images [article]

Vishal Asnani, Xi Yin, Tal Hassner, Xiaoming Liu
2021 arXiv   pre-print
State-of-the-art (SOTA) Generative Models (GMs) can synthesize photo-realistic images that are hard for humans to distinguish from genuine photos.  ...  We define a novel problem, "model parsing", as estimating GM network architectures and training loss functions by examining their generated images – a task seemingly impossible for human beings.  ...  Fig. 3 3 : t SNE visualization for ground-truth feature vectors for (a) network architecture, (b) fine-level loss function and (c) network architecture and fine-level loss function combined.  ... 
arXiv:2106.07873v1 fatcat:ehlstw55d5cwhd5fcvre44zc7a
« Previous Showing results 1 — 15 out of 167 results