Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Optimal denoising in redundant representations.

Martin Raphan1, Eero P Simoncelli

  • 1Howard Hughes Medical Institute, Center for Neural Science, New York, NY 10003, USA. raphan@cims.nyu.edu

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|July 18, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Responses of neurons in macaque V4 to object and texture images.

bioRxiv : the preprint server for biology·2025
Same author

Detection of moving objects using self-motion constraints on optic flow.

ArXiv·2025
Same author

Contrastive-Equivariant Self-Supervised Learning Improves Alignment with Primate Visual Area IT.

Advances in neural information processing systems·2025
Same author

Perceptual learning improves discrimination but does not reduce distortions in appearance.

PLoS computational biology·2025
Same author

Fixational eye movements enhance the precision of visual information transmitted by the primate retina.

Nature communications·2024
Same author

Neuronal and Behavioral Responses to Naturalistic Texture Images in Macaque Monkeys.

The Journal of neuroscience : the official journal of the Society for Neuroscience·2024
Same journal

Change-Prior-Guided Unsupervised Change Detection of Heterogeneous Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

AgonicDreamer: Enhancing Multi-View Consistency in Text-to-3D Generation via Rectified Score Distillation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

BiCM-Prompt: Bidirectional Cross-Modal Prompt Tuning for Class-Incremental Learning on Multisource Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

GoP-based Quality Enhancement on Video Compression.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Align then Tensorize: Multi-Level Consistent Anchor Graph Learning for Scalable Multi-View Clustering.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Beyond Fidelity: Diverse Image Synthesis via Retrieval-Augmented Diffusion.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
See all related articles

Overcomplete image representations improve denoising by reducing expected image-domain mean-squared error (MSE). New methods optimize estimators for image-domain MSE, significantly outperforming subband-domain optimization.

Area of Science:

  • Digital Image Processing
  • Signal Processing
  • Computational Imaging

Background:

  • Image denoising methods often target mean-squared error (MSE) within multiscale decomposition subbands.
  • High-quality denoising frequently utilizes overcomplete representations, but subband MSE minimization doesn't guarantee optimal image-domain MSE.

Purpose of the Study:

  • To prove that overcomplete representations, even with suboptimal subband MSE minimization, yield better image-domain MSE.
  • To develop a method for jointly optimizing subband estimators for image-domain MSE using overcomplete representations.
  • To demonstrate the effectiveness of this new optimization approach through simulations.

Main Methods:

  • Theoretical proof showing expected image-domain MSE is minimized with redundant spatial replication (e.g., cycle spinning).

Related Experiment Videos

  • Development of an extended Stein's Unbiased Risk Estimate (SURE) for adaptive, image-domain MSE optimization.
  • Application of novel estimators based on linear combinations of localized bump functions.

    • Main Results:

    • Overcomplete representations with spatial replication reduce expected image-domain MSE compared to nonredundant ones.
    • Joint optimization for image-domain MSE significantly enhances denoising performance.
    • The proposed adaptive SURE-based optimization method achieves substantial performance gains.

    Conclusions:

    • Optimizing image denoising estimators for image-domain MSE using overcomplete representations offers superior results.
    • The developed adaptive SURE method effectively exploits overcompleteness for improved denoising.
    • This approach significantly outperforms traditional subband-optimized or nonredundant representation methods.