On the Convergence of PatchMatch and Its Variants

Ehret, Thibaud; Arias, Pablo

doi:10.1109/cvpr.2018.00123

Cited by 2 publications

(2 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, other global patch search methods, such as PatchMatch [5], can also be used. While PatchMatch has more convergence guarantees, see [22], it is also less computationally efficient. PatchMatch is very efficient when looking for very few (namely one or two) nearest neighbors but struggles for a large number of nearest neighbors.…”

Section: Self-similarity Background Modelingmentioning

confidence: 99%

How to Reduce Anomaly Detection in Images to Anomaly Detection in Noise

Ehret¹,

Davy²,

Delbracio³

et al. 2019

Image Processing on Line

Self Cite

View full text Add to dashboard Cite

Anomaly detectors address the difficult problem of detecting automatically exceptions in a background image, that can be as diverse as a fabric or a mammography. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by a simple noise model, and allows the calculation of rigorous detection thresholds. Our approach is therefore unsupervised and works on arbitrary images. The residual images can favorably be computed on dense features of neural networks. Our detector is powered by the a contrario detection theory, which avoids over-detection by fixing detection thresholds taking into account the multiple tests. Source Code The reviewed source code and documentation for this algorithm are available from the web page of this article 1. Compilation and usage instruction are included in the README.txt file of the archive.

show abstract

Section: Self-similarity Background Modelingmentioning

confidence: 99%

How to Reduce Anomaly Detection in Images to Anomaly Detection in Noise

Ehret¹,

Davy²,

Delbracio³

et al. 2019

Image Processing on Line

Self Cite

View full text Add to dashboard Cite

show abstract

“…The convergence of this matching algorithm has been proved (for the original one as well as for a modified version like this one) using the framework presented in [7].…”

Section: Generalized Patchmatchmentioning

confidence: 99%

Automatic Detection of Internal Copy-Move Forgeries in Images

Ehret¹

2018

Image Processing on Line

Self Cite

View full text Add to dashboard Cite

This article presents an implementation and discussion of the recently proposed 'Efficient Dense-Field Copy-Move Forgery Detection' by Cozzolino et al. This method is a forgery detection based on a dense field of descriptors chosen to be invariant by rotation. Zernike moments were suggested in the original article. An efficient matching of the descriptors is then performed using PatchMatch, which is extremely efficient to find duplicate regions. Regions matched by PatchMatch are processed to find the final detections. This allows a precise and accurate detection of copy-move forgeries inside a single suspicious image. We also extend successfully the method to the use of dense SIFT descriptors and show that they are better at detecting forgeries using Poisson editing. Source Code The reviewed source code and documentation for this algorithm are available from the web page of this article 1. Compilation and usage instruction are included in the README.txt file of the archive.

show abstract

On the Convergence of PatchMatch and Its Variants

Cited by 2 publications

References 18 publications

How to Reduce Anomaly Detection in Images to Anomaly Detection in Noise

How to Reduce Anomaly Detection in Images to Anomaly Detection in Noise

Automatic Detection of Internal Copy-Move Forgeries in Images

Contact Info

Product

Resources

About