Towards a Visual Privacy Advisor: Understanding and Predicting Privacy Risks in Images
Not on the map. Why?
Abstract (arXiv)
With an increasing number of users sharing information online, privacy implications entailing such actions are a major concern. For explicit content, such as user profile or GPS data, devices (e.g. mobile phones) as well as web services (e.g. Facebook) offer to set privacy settings in order to enforce the users' privacy preferences. We propose the first approach that extends this concept to image content in the spirit of a Visual Privacy Advisor. First, we categorize personal information in images into 68 image attributes and collect a dataset, which allows us to train models that predict such information directly from images. Second, we run a user study to understand the privacy preferences of different users w.r.t. such attributes. Third, we propose models that predict user specific privacy score from images in order to enforce the users' privacy preferences. Our model is trained to predict the user specific privacy risk and even outperforms the judgment of the users, who often fail to follow their own privacy preferences on image data.
Key facts. Cited by 101 papers in the map (data through 7 Oct 2026). First posted on arXiv 30 Mar 2017.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2021: 12; 2022: 24; 2023: 17; 2024: 13; 2025: 22; 2026: 13.
101 papers.
Sort:
By area:
- 2021VisualGPT: Data-efficient Adaptation of Pretrained Language Models for Image Captioningcited by 190
- 2021From Show to Tell: A Survey on Deep Learning-based Image Captioningcited by 137
- 2021A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learningcited by 82
- 2023Positive-Augmented Contrastive Learning for Image and Video Captioning Evaluationcited by 78
- 2023Rank2Tell: A Multimodal Driving Dataset for Joint Importance Ranking and Reasoningcited by 63
- 2022SPAct: Self-supervised Privacy Preservation for Action Recognitioncited by 47
- 2022Deep Learning Approaches on Image Captioning: A Reviewcited by 43
- 2021Generate, Prune, Select: A Pipeline for Counterspeech Generation against Online Hate Speechcited by 42
- 2021Human-like Controllable Image Captioning with Verb-specific Semantic Rolescited by 41
- 2023Image Shortcut Squeezing: Countering Perturbative Availability Poisons with Compressioncited by 39
- 2022Gradient Obfuscation Gives a False Sense of Security in Federated Learningcited by 37
- 2022Measuring Representational Harms in Image Captioningcited by 36
38 more
- 2021Probabilistic Modeling of Semantic Ambiguity for Scene Graph Generationcited by 35
- 2023HowToCaption: Prompting LLMs to Transform Video Annotations at Scalecited by 34
- 2023Multi-modal Machine Learning in Engineering Design: A Review and Future Directionscited by 27
- 2022VLTinT: Visual-Linguistic Transformer-in-Transformer for Coherent Video Paragraph Captioningcited by 25
- 2023Movie101: A New Movie Understanding Benchmarkcited by 25
- 2023TeD-SPAD: Temporal Distinctiveness for Self-supervised Privacy-preservation for video Anomaly Detectioncited by 23
- 2021Towards Diverse Paragraph Captioning for Untrimmed Videoscited by 22
- 2023IC3: Image Captioning by Committee Consensuscited by 20
- 2022VLCap: Vision-Language with Contrastive Learning for Coherent Video Paragraph Captioningcited by 19
- 2022Affection: Learning Affective Explanations for Real-World Visual Datacited by 19
- 2022PrivHAR: Recognizing Human Actions From Privacy-preserving Lenscited by 16
- 2022Privacy-Preserving Action Recognition via Motion Difference Quantizationcited by 14
- 2022Learning Distinct and Representative Styles for Image Captioningcited by 12
- 2024Privacy-Aware Visual Language Modelscited by 12
- 2022CapOnImage: Context-driven Dense-Captioning on Imagecited by 11
- 2023Simple Token-Level Confidence Improves Caption Correctnesscited by 11
- 2021Going Grayscale: The Road to Understanding and Improving Unlearnable Examplescited by 10
- 2022Show, Interpret and Tell: Entity-aware Contextualised Image Captioning in Wikipediacited by 10
- 2022Rethinking the Reference-based Distinctive Image Captioningcited by 8
- 2021Discriminative Latent Semantic Graph for Video Captioningcited by 7
- 2022On Distinctive Image Captioning via Comparing and Reweightingcited by 7
- 2022Distinctive Image Captioning via CLIP Guided Group Optimizationcited by 6
- 2023Modeling the Trade-off of Privacy Preservation and Activity Recognition on Low-Resolution Imagescited by 6
- 2023Free-ATM: Exploring Unsupervised Learning on Diffusion-Generated Images with Free Attention Maskscited by 6
- 2022MaSS: Multi-attribute Selective Suppressioncited by 5
- 2022CapEnrich: Enriching Caption Semantics for Web Images via Cross-modal Pre-trained Knowledgecited by 5
- 2022Contextual Explainable Video Representation: Human Perception-based Understandingcited by 5
- 2024BIV-Priv-Seg: Locating Private Content in Images Taken by People With Visual Impairmentscited by 5
- 2025REVAL: A Comprehension Evaluation on Reliability and Values of Large Vision-Language Modelscited by 5
- 2025Evaluation of Human Visual Privacy Protection: A Three-Dimensional Framework and Benchmark Datasetcited by 5
- 2022DRAG: Dynamic Region-Aware GCN for Privacy-Leaking Image Detectioncited by 4
- 2022Content-based Graph Privacy Advisorcited by 4
- 2023Improving Reference-based Distinctive Image Captioning with Contrastive Rewardscited by 4
- 2024Image-guided topic modeling for interpretable privacy classificationcited by 4
- 2024FG-CXR: A Radiologist-Aligned Gaze Dataset for Enhancing Interpretability in Chest X-Ray Report Generationcited by 4
- 2025EgoPrivacy: What Your First-Person Camera Says About You?cited by 4
- 2025Unveiling Trust in Multimodal Large Language Models: Evaluation, Analysis, and Mitigationcited by 4
- 2025Privy: Envisioning and Mitigating Privacy Risks for Consumer-facing AI Product Conceptscited by 4
50 most cited of 101 All 101 (JSON)
- 2026SpatialTrust: A Benchmark for Environmental Risk Recognition in Secure Authentication30 Aug 2026
- 2026Face De-Identification: A Domain-Centric Survey from Capture to Processing28 Jul 2026
- 2026CAPED: Context-Aware Privacy Exposure Defense for Mobile GUI Agents10 Jun 2026
- 2026Seeing Without Exposing: Adaptive Privacy Control for Open-World, Context-Hungry MLLMs5 Jun 2026
- 2026When Recovery Matters: The Blind Spot of Surrogate Privacy in MLLM Editing5 Jun 2026
- 2026Unsafe2Safe: Controllable Image Anonymization for Downstream Utility30 Mar 2026
- 2026Towards Context-Aware Image Anonymization with Multi-Agent Reasoning29 Mar 2026
- 2026Rethinking Visual Privacy: A Compositional Privacy Risk Framework for Severity Assessment with VLMs23 Mar 2026
- 2026When Visual Privacy Protection Meets Multimodal Large Language Models14 Mar 2026
- 2026Neural Gate: Mitigating Privacy Risks in LVLMs via Neuron-Level Gradient Gating13 Mar 2026
- 2026Which private attributes do VLMs agree on and predict well?8 Feb 2026
- 2026PrivLEX: Detecting legal concepts in images through Vision-Language Models14 Jan 2026
38 more
- 2026MPCI-Bench: A Benchmark for Multimodal Pairwise Contextual Integrity Evaluation of Language Model Agents13 Jan 2026
- 2025Cross-modal Counterfactual Explanations: Uncovering Decision Factors and Dataset Biases in Subjective Classification21 Dec 2025
- 2025StegaVAR: Privacy-Preserving Video Action Recognition via Steganographic Domain Analysis14 Dec 2025
- 2025A Distributed Framework for Privacy-Enhanced Vision Transformers on the Edge10 Dec 2025
- 2025When Privacy Meets Recovery: The Overlooked Half of Surrogate-Driven Privacy Preservation for MLLM Editing8 Dec 2025
- 2025GEO-Detective: Unveiling Location Privacy Risks in Images with LLM Agents27 Nov 2025
- 2025Privacy Beyond Pixels: Latent Anonymization for Privacy-Preserving Video Understanding11 Nov 2025
- 2025Merlin L48 Spectrogram Dataset31 Oct 2025
- 2025From See to Shield: ML-Assisted Fine-Grained Access Control for Visual Data22 Oct 2025
- 2025On the Robustness of Vision-Language Models in Zero-shot Privacy Classification10 Oct 2025
- 2025The impact of abstract and object tags on image privacy classification9 Oct 2025
- 2025Privy: Envisioning and Mitigating Privacy Risks for Consumer-facing AI Product Concepts27 Sep 2025
- 2025Automatic Identification and Description of Jewelry Through Computer Vision and Neural Networks for Translators and Interpreters31 Aug 2025
- 2025Unveiling Trust in Multimodal Large Language Models: Evaluation, Analysis, and Mitigation21 Aug 2025
- 2025Beyond Blanket Masking: Examining Granularity for Privacy Protection in Images Captured by Blind and Low Vision Users12 Aug 2025
- 2025Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models11 Aug 2025
- 2025Evaluation of Human Visual Privacy Protection: A Three-Dimensional Framework and Benchmark Dataset18 Jul 2025
- 2025EgoPrivacy: What Your First-Person Camera Says About You?13 Jun 2025
- 2025REVAL: A Comprehension Evaluation on Reliability and Values of Large Vision-Language Models20 Mar 2025
- 2025Learning Privacy from Visual Entities16 Mar 2025
- 2025RONA: Pragmatically Diverse Image Captioning with Coherence Relations14 Mar 2025
- 2025Video-DPRP: A Differentially Private Approach for Visual Privacy-Preserving Video Human Activity Recognition3 Mar 2025
- 2025Beyond RNNs: Benchmarking Attention-Based Image Captioning Models26 Feb 2025
- 2024Multi-PA: A Multi-perspective Benchmark on Privacy Assessment for Large Vision-Language Models27 Dec 2024
- 2024Intermediate Outputs Are More Sensitive Than You Think1 Dec 2024
- 2024FG-CXR: A Radiologist-Aligned Gaze Dataset for Enhancing Interpretability in Chest X-Ray Report Generation23 Nov 2024
- 2024Low-Latency Video Anonymization for Crowd Anomaly Detection: Privacy Versus Performance24 Oct 2024
- 2024GEM-VPC: A dual Graph-Enhanced Multimodal integration for Video Paragraph Captioning12 Oct 2024
- 2024Image-guided topic modeling for interpretable privacy classification27 Sep 2024
- 2024BIV-Priv-Seg: Locating Private Content in Images Taken by People With Visual Impairments25 Jul 2024
- 2024Privacy-Aware Visual Language Models27 May 2024
- 2024Explaining models relating objects and privacy2 May 2024
- 2024AIGeN: An Adversarial Approach for Instruction Generation in VLN15 Apr 2024
- 2024Deep Privacy Funnel Model: From a Discriminative to a Generative Approach with an Application to Face Recognition3 Apr 2024
- 2024A Self-supervised Pressure Map human keypoint Detection Approch: Optimizing Generalization and Computational Efficiency Across Datasets22 Feb 2024
- 2024Multi-modal News Understanding with Professionally Labelled Videos (ReutersViLNews)23 Jan 2024
- 2023CLearViD: Curriculum Learning for Video Description8 Nov 2023
- 2023On Responsible Machine Learning Datasets with Fairness, Privacy, and Regulatory Norms24 Oct 2023
50 newest of 101 All 101 (JSON)
- 2022Deep Learning Approaches on Image Captioning: A Reviewcited by 43
- 2024Privacy-Aware Visual Language Modelscited by 12
- 2025REVAL: A Comprehension Evaluation on Reliability and Values of Large Vision-Language Modelscited by 5
- 2025Unveiling Trust in Multimodal Large Language Models: Evaluation, Analysis, and Mitigationcited by 4
- 2024Multi-PA: A Multi-perspective Benchmark on Privacy Assessment for Large Vision-Language Modelscited by 3
- 2025Beyond Blanket Masking: Examining Granularity for Privacy Protection in Images Captured by Blind and Low Vision Userscited by 2
- 2025Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Modelscited by 1
- 2025On the Robustness of Vision-Language Models in Zero-shot Privacy Classificationcited by 1
- 2025GEO-Detective: Unveiling Location Privacy Risks in Images with LLM Agentscited by 0
- 2026Which private attributes do VLMs agree on and predict well?cited by 0
- 2026Neural Gate: Mitigating Privacy Risks in LVLMs via Neuron-Level Gradient Gatingcited by 0
- 2026When Visual Privacy Protection Meets Multimodal Large Language Modelscited by 0
12 most cited of 14 All 14 (JSON)
- 2026Seeing Without Exposing: Adaptive Privacy Control for Open-World, Context-Hungry MLLMs5 Jun 2026
- 2026When Recovery Matters: The Blind Spot of Surrogate Privacy in MLLM Editing5 Jun 2026
- 2026When Visual Privacy Protection Meets Multimodal Large Language Models14 Mar 2026
- 2026Neural Gate: Mitigating Privacy Risks in LVLMs via Neuron-Level Gradient Gating13 Mar 2026
- 2026Which private attributes do VLMs agree on and predict well?8 Feb 2026
- 2025GEO-Detective: Unveiling Location Privacy Risks in Images with LLM Agents27 Nov 2025
- 2025On the Robustness of Vision-Language Models in Zero-shot Privacy Classification10 Oct 2025
- 2025Unveiling Trust in Multimodal Large Language Models: Evaluation, Analysis, and Mitigation21 Aug 2025
- 2025Beyond Blanket Masking: Examining Granularity for Privacy Protection in Images Captured by Blind and Low Vision Users12 Aug 2025
- 2025Through Their Eyes: User Perceptions on Sensitive Attribute Inference of Social Media Videos by Visual Language Models11 Aug 2025
- 2025REVAL: A Comprehension Evaluation on Reliability and Values of Large Vision-Language Models20 Mar 2025
- 2024Multi-PA: A Multi-perspective Benchmark on Privacy Assessment for Large Vision-Language Models27 Dec 2024
12 newest of 14 All 14 (JSON)
- 2021VisualGPT: Data-efficient Adaptation of Pretrained Language Models for Image Captioningcited by 190
- 2021From Show to Tell: A Survey on Deep Learning-based Image Captioningcited by 137
- 2021A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learningcited by 82
- 2023Positive-Augmented Contrastive Learning for Image and Video Captioning Evaluationcited by 78
- 2023Rank2Tell: A Multimodal Driving Dataset for Joint Importance Ranking and Reasoningcited by 63
- 2022SPAct: Self-supervised Privacy Preservation for Action Recognitioncited by 47
- 2021Generate, Prune, Select: A Pipeline for Counterspeech Generation against Online Hate Speechcited by 42
- 2021Human-like Controllable Image Captioning with Verb-specific Semantic Rolescited by 41
- 2023Image Shortcut Squeezing: Countering Perturbative Availability Poisons with Compressioncited by 39
- 2022Gradient Obfuscation Gives a False Sense of Security in Federated Learningcited by 37
- 2022Measuring Representational Harms in Image Captioningcited by 36
- 2021Probabilistic Modeling of Semantic Ambiguity for Scene Graph Generationcited by 35
12 most cited of 85 All 85 (JSON)
- 2026SpatialTrust: A Benchmark for Environmental Risk Recognition in Secure Authentication30 Aug 2026
- 2026Face De-Identification: A Domain-Centric Survey from Capture to Processing28 Jul 2026
- 2026CAPED: Context-Aware Privacy Exposure Defense for Mobile GUI Agents10 Jun 2026
- 2026Unsafe2Safe: Controllable Image Anonymization for Downstream Utility30 Mar 2026
- 2026Towards Context-Aware Image Anonymization with Multi-Agent Reasoning29 Mar 2026
- 2026Rethinking Visual Privacy: A Compositional Privacy Risk Framework for Severity Assessment with VLMs23 Mar 2026
- 2026PrivLEX: Detecting legal concepts in images through Vision-Language Models14 Jan 2026
- 2025Cross-modal Counterfactual Explanations: Uncovering Decision Factors and Dataset Biases in Subjective Classification21 Dec 2025
- 2025StegaVAR: Privacy-Preserving Video Action Recognition via Steganographic Domain Analysis14 Dec 2025
- 2025A Distributed Framework for Privacy-Enhanced Vision Transformers on the Edge10 Dec 2025
- 2025When Privacy Meets Recovery: The Overlooked Half of Surrogate-Driven Privacy Preservation for MLLM Editing8 Dec 2025
- 2025Privacy Beyond Pixels: Latent Anonymization for Privacy-Preserving Video Understanding11 Nov 2025
12 newest of 85 All 85 (JSON)