91,496 Hits in 5.6 sec

On the scaling of polynomial features for representation matching [article]

Siddhartha Brahma
2018 arXiv   pre-print
Using the natural language inference task as an example, we investigate the use of scaled polynomials of degree 2 and above as matching features.  ...  In many neural models, new features as polynomial functions of existing ones are used to augment representations.  ...  CONCLUSION AND FUTURE DIRECTIONS In this paper, we explore the use of higher degree polynomial features and scaling to derive new features when two distributed representations have to be matched or compared  ... 
arXiv:1802.07374v1 fatcat:fa5z4iidifg4bitrkqfaf3pvpm

HERS: Homomorphically Encrypted Representation Search [article]

Joshua J. Engelsma and Anil K. Jain and Vishnu Naresh Boddeti
2022 arXiv   pre-print
scale (500 seconds; 275× speed up over state-of-the-art for encrypted search against a gallery of 100 million).  ...  Numerical results on large galleries of face, fingerprint, and object datasets such as ImageNet show that, for the first time, accurate and fast image search within the encrypted domain is feasible at  ...  ACKNOWLEDGEMENTS This research was supported in part by a grant from The NSF Center for Identification Technology Research (CITeR).  ... 
arXiv:2003.12197v3 fatcat:ffjuaoil5valtovrsk5jawglmq

HEFT: Homomorphically Encrypted Fusion of Biometric Templates [article]

Luke Sperling, Nalini Ratha, Arun Ross, Vishnu Naresh Boddeti
2022 arXiv   pre-print
Experimental evaluation for template fusion and matching of face and voice biometrics shows that HEFT (i) improves biometric verification performance by 11.07 representations while compressing the feature  ...  Given a pair of encrypted feature vectors, we perform the following ciphertext operations, i) feature concatenation, ii) fusion and dimensionality reduction through a learned linear projection, iii) scale  ...  Next, we outline two matrix-vector multiplication techniques, one that is better suited for small-scale datasets and the other for large-scale datasets.  ... 
arXiv:2208.07241v1 fatcat:ss53ip4xxzhrrgylpn7echlvpe

Asymmetric Feature Maps with Application to Sketch Based Retrieval [article]

Giorgos Tolias, Ondřej Chum
2017 arXiv   pre-print
To demonstrate the advantages of the AFM method, we derive a short vector image representation that, due to asymmetric feature maps, supports efficient scale and translation invariant sketch-based image  ...  The results are boosted by an image-based average query expansion, exceeding significantly the state of the art on standard benchmarks.  ...  No additional storage is required on the database side to evaluate the kernel. The same holds for efficient match kernels, as (1) is a normalized sum of feature maps.  ... 
arXiv:1704.03946v1 fatcat:z4miwdqu2rambngexqh46oxyni

Multiscale object representation using surface patches

Wassim Alami, Gregory Dudek, David P. Casasent
1994 Intelligent Robots and Computer Vision XIII: Algorithms and Computer Vision  
The extraction of simple uniform curvature features is limited by the fact that the optimal scale of processing for a single object is very di cult to determine.  ...  The representation is based on surface patches with uniform curvature properties extracted at multiple scales.  ...  ACKNOWLEDGEMENTS The authors gratefully acknowledge the nancial support of the Natural Sciences and Engineering Research Council and the Canadian Federal Centres of Excellence Program.  ... 
doi:10.1117/12.188884 fatcat:6hmtbrqpjzbopkzlok7ac3ttge

Image Local Features Description through Polynomial Approximation

Fawad, Muhibur Rahman, Muhammad Jamil Khan, Muhammad Adeel Asghar, Yasar Amin, Salman Badnava, Seyed Sajad Mirjavadi
2019 IEEE Access  
The correlation coefficient is used instead of Euclidean distance to improve the matching accuracy.  ...  The proposed descriptor incorporate polynomials of various degrees to approximate the local patch within the image.  ...  The influence of parameters such as patch radius, matching threshold, and the degree of the polynomial, on the feature matching performance, is examined.  ... 
doi:10.1109/access.2019.2959326 fatcat:fklqgmfa75egzjxrhggtdy5nmq

A Resilient Image Matching Method with an Affine Invariant Feature Detector and Descriptor [article]

Biao Zhao, Shigang Yue
2017 arXiv   pre-print
With systematic experiments, we can prove that the proposed method of feature detector and descriptor outperforms other state-of-the-art feature matching algorithms especially on view points robustness  ...  Image feature matching is to seek, localize and identify the similarities across the images. The matched local features between different images can indicate the similarities of their content.  ...  Acknowledgements We would like to express our gratitude to the projects EU FP7 LIVECODE(295151), EU FP7 HAZCEPT(318907), EUHorizon 2020 STEP2DYNA(691154), EUHorizon 2020 EN-RICHME(643691) for financial  ... 
arXiv:1802.09623v1 fatcat:ptb6uujzhvclhdmb7e5ewknlri

A data-driven approach to feature space selection for robust micro-endoscopic image reconstruction

Pascal Bourdon, David Helbert
2017 2017 IEEE International Conference on Image Processing (ICIP)  
In the article we propose a new on-line feature space selection strategy for displacement field estimation in the context of multi-view reconstruction of biological images acquired by a multi-photon micro-endoscope  ...  methods, we will show how on-line learning combined with a classical method such as Digital Image Correlation (DIC) can contribute to the improvement of convex optimization-based template matching techniques  ...  projection onto a new representation domain called feature space f(I ), in the case of feature matching.  ... 
doi:10.1109/icip.2017.8296680 dblp:conf/icip/BourdonH17 fatcat:dmbu72tzvjg6pbfjjtayxarnea

Shape representation by multiscale contour approximation

A. Bengtsson, J.-O. Eklundh
1991 IEEE Transactions on Pattern Analysis and Machine Intelligence  
In this approach one considers a one-parameter family of representations of a shape, ihe parameter being scale.  ...  For each inflexion point i at scale kf + 1 that has not been matched exactly, we set up a list of matching candidates from level kf.  ... 
doi:10.1109/34.67634 fatcat:zvkawsfqbrbdzod65akcbki4ne

Affine Normalized Krawtchouk Moments Based Face Recognition

B.H. Shekar, D.S. Rajesh
2015 Procedia Computer Science  
Experimental results on the ORL dataset and a subset of pose and illumination variant FERET dataset have shown the classification capability of our descriptor for face recognition applications.  ...  By using the sparse representation concept, classification of face images is done.  ...  (of 5 scales and 8 orientations) of the face image to remove redundancy during feature representation and used sparse representation based classification for face recognition.  ... 
doi:10.1016/j.procs.2015.08.014 fatcat:en572rgwhzgnbmm6hwmwuh2jte

Signature based Fuzzy Vaults with Boosted Feature Selection

George S. Eskander, Robert Sabourin, Eric Granger
2011 2011 IEEE Workshop on Computational Intelligence in Biometrics and Identity Management (CIBIM)  
In this context, bio-cryptography systems based on the handwritten signatures may be considered for enhance security.  ...  The indexes of selected features correspond to the most stable and discriminant features from a user's signature images, and are used to encode user-specific FVs.  ...  of the 256 features in this scale representation), and WD-FV using the multi-scale WD-ESC feature vector.  ... 
doi:10.1109/cibim.2011.5949215 dblp:conf/cibim/EskanderSG11 fatcat:zta3ldu5fnbm7acuvs44d7temy

Offline Signature-Based Fuzzy Vault (OSFV: Review and New Results [article]

George S. Eskander, Robert Sabourin, Eric Granger
2014 arXiv   pre-print
In this system, a machine learning approach based on the dissimilarity representation concept is employed to select a reliable feature representation adapted for the fuzzy vault scheme.  ...  Some variants of this system are proposed for enhanced accuracy and security. In particular, a new method that adapts user key size is presented.  ...  This process results in a matching set (Ā,P ) = ((B Ã ), p ← (B Ã )), where p ← (B Ã ) represents the projection of the matching features on the polynomial space. Chaff filtering is done as follows.  ... 
arXiv:1408.3985v1 fatcat:7xvxklp52jcqdoarfudo4kln6y

Hermite deformable contours

D.M. Gavrila
1996 Proceedings of 13th International Conference on Pattern Recognition  
Our approach t o c o n tour tracking decouples the e ects of transformation and deformation, using a template matching strategy to robustly account for the transformation e ect.  ...  The Hermite representation allows a compact representation of curved shapes, without the smoothing out of corners. It is also well suited for both interactive and tracking applications.  ...  Acknowledgements The author thanks Larry Davis for his continued support.  ... 
doi:10.1109/icpr.1996.546005 dblp:conf/icpr/Gavrila96 fatcat:a2lrsdyx7jd4bnoi547x45tlum

Content-based shape retrieval using different shape descriptors: a comparative study

Dengsheng Zhang, Guojun Lu
2001 IEEE International Conference on Multimedia and Expo, 2001. ICME 2001.  
2 Outline • Motivations • Comparison of Contour Descriptors o Fourier Descriptors (FD) o Curvature Scale Space Descriptors (CSSD) o Retrieval Experiments o Discussions • Comparison of Region Descriptors  ...  GD is a more intuitive shape representation than ZMD. o CSSD is not suitable for efficient indexing due to the expensive matching and variation of feature dimensions o Based on these facts, we recommend  ...  ZMD-I • Zernike Polynomials: • Feature domains.  ... 
doi:10.1109/icme.2001.1237928 dblp:conf/icmcs/ZhangL01 fatcat:kmxozl2kzbdptlyigj3v6f54im

Polynomial transformation model for frame-to-frame registration in an adaptive optics confocal scanning laser ophthalmoscope

Hao Chen, Yi He, Ling Wei, Jinsheng Yang, Xiqi Li, Guohua Shi, Yudong Zhang
2019 Biomedical Optics Express  
motions such as scale changes, shearing and rotation motions, and so on.  ...  Then we generated the polynomial transformation model and provided its close-form solution for consecutively frame-to-frame AOSLO retina image registration, allowing one to consider more general retinal  ...  Acknowledgments We thank LetPub ( for its linguistic assistance during the preparation of this manuscript.  ... 
doi:10.1364/boe.10.004589 pmid:31565511 pmcid:PMC6757461 fatcat:nexf2eov2nekbezg6pnhcwaam4
« Previous Showing results 1 — 15 out of 91,496 results