mlsanityData to 14 Oct 2026

LDP-Feat: Image Features with Local Differential Privacy

arXiv 2308.11223 ↗First posted 22 Aug 2023cs.CV cs.CRcited by 10

Not on the map. Why?

Abstract (arXiv)

Modern computer vision services often require users to share raw feature descriptors with an untrusted server. This presents an inherent privacy risk, as raw descriptors may be used to recover the source images from which they were extracted. To address this issue, researchers recently proposed privatizing image features by embedding them within an affine subspace containing the original feature as well as adversarial feature samples. In this paper, we propose two novel inversion attacks to show that it is possible to (approximately) recover the original image features from these embeddings, allowing us to recover privacy-critical image content. In light of such successes and the lack of theoretical privacy guarantees afforded by existing visual privacy methods, we further propose the first method to privatize image features via local differential privacy, which, unlike prior approaches, provides a guaranteed bound for privacy leakage regardless of the strength of the attacks. In addition, our method yields strong performance in visual localization as a downstream task while enjoying the privacy guarantee.

Key facts. Cited by 10 papers in the map (data through 14 Oct 2026). Builds on 26 of its 53 references that are in the map. First posted on arXiv 22 Aug 2023.

Builds on

26 of 53 references.

References · 26

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2013Visualizing and Understanding Convolutional Networkscited by 2,290
  3. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 737
  4. 2014Understanding Deep Image Representations by Inverting Themcited by 403
  5. 2015Understanding Neural Networks Through Deep Visualizationcited by 334
  6. 2016Generating Images with Perceptual Similarity Metrics based on Deep Networkscited by 222
  7. 2017Working hard to know your neighbor's margins: Local descriptor learning losscited by 212
  8. 2019Learning with Average Precision: Training Image Retrieval with a Listwise Losscited by 185
  9. 2015Inverting Visual Representations with Convolutional Networkscited by 170
  10. 2019Revealing Scenes by Inverting Structure from Motion Reconstructionscited by 66
  11. 2016Examining the Impact of Blur on Recognition by Convolutional Networkscited by 58
  12. 2018Towards Privacy-Preserving Visual Recognition via Adversarial Training: A Pilot Studycited by 48
14 more
  1. 2019Privacy Preserving Image-Based Localizationcited by 47
  2. 2016Defeating Image Obfuscation with Deep Learningcited by 43
  3. 2018Local Descriptors Optimized for Average Precisioncited by 43
  4. 2020Local Differential Privacy and Its Applications: A Comprehensive Surveycited by 41
  5. 2019Adversarial Learning of Privacy-Preserving and Task-Oriented Representationscited by 36
  6. 2014Differential Privacy: An Economic Method for Choosing Epsiloncited by 35
  7. 2018Learning Privacy Preserving Encodings through Adversarial Trainingcited by 30
  8. 2020A Comprehensive Survey on Local Differential Privacy Toward Data Statistics and Analysiscited by 26
  9. 2018Differential Privacy Techniques for Cyber Physical Systems: A Surveycited by 21
  10. 2021NinjaDesc: Content-Concealing Visual Descriptors via Adversarial Learningcited by 21
  11. 2021How Privacy-Preserving are Line Clouds? Recovering Scene Details from 3D Linescited by 19
  12. 2020Privacy Preserving Visual SLAMcited by 15
  13. 2015Image Reconstruction from Bag-of-Visual-Wordscited by 6
  14. 2019Privacy-Preserving Deep Action Recognition: An Adversarial Learning Framework and A New Datasetcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 3; 2025: 2; 2026: 5.

10 papers.

  1. 2025Gaussian Splatting Feature Fields for Privacy-Preserving Visual Localizationcited by 10
  2. 2024Thinking Racial Bias in Fair Forgery Detection: Models, Datasets and Evaluationscited by 8
  3. 2024Obfuscation Based Privacy Preserving Representations are Recoverable Using Neighborhood Informationcited by 6
  4. 2025LeakyCLIP: Extracting Training Data from CLIPcited by 4
  5. 2026LDP-Slicing: Local Differential Privacy for Images via Randomized Bit-Plane Slicingcited by 2
  6. 2024Calibrating Practical Privacy Risks for Differentially Private Machine Learningcited by 0
  7. 2026Privacy-Preserving Structureless Visual Localization via Image Obfuscationcited by 0
  8. 2026Revisiting Geometric Obfuscation with Dual Convergent Lines for Privacy-Preserving Image Queries in Visual Localizationcited by 0
  9. 2026Seeing Through the Weights: Privacy Leakage in Scene Coordinate Regressioncited by 0
  10. 2026Misanthrope: A Privacy-Preserving Keypoint Detectorcited by 0