Privacy-Preserving Semantic Segmentation from Ultra-Low-Resolution RGB Inputs

2025/07/21 by Xuying Huang, Huang, Xuying, Sicong Pan +7
Computer Science · Engineering · #Advanced Neural Network Applications #Adversarial Robustness in Machine Learning #Autonomous Vehicle Technology and Safety #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Robotics (cs.RO)

paper · pdf · doi:10.48550/arxiv.2507.16034

openalex publication_date 2025/07/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/30

Abstract

RGB-based semantic segmentation has become a mainstream approach for visual perception and is widely applied in a variety of downstream tasks. However, existing methods typically rely on high-resolution RGB inputs, which may expose sensitive visual content in privacy-critical environments. Ultra-low-resolution RGB sensing suppresses sensitive information directly during image acquisition, making it an attractive privacy-preserving alternative. Nevertheless, recovering semantic segmentation from ultra-low-resolution RGB inputs remains highly challenging due to severe visual degradation. In this work, we introduce a novel fully joint-learning framework to mitigate the optimization conflicts exacerbated by visual degradation for ultra-low-resolution semantic segmentation. Experiments demonstrate that our method outperforms representative baselines in semantic segmentation performance and our ultra-low-resolution RGB input achieves a favorable trade-off between privacy preservation and semantic segmentation performance. We deploy our privacy-preserving semantic segmentation method in a real-world robotic object-goal navigation task, demonstrating successful downstream task execution even under severe visual degradation.

Citations

• SafeTriage: Facial Video De-identification for Privacy-Preserving Stroke Triage
• Privacy Risks of Robot Vision: A User Study on Image Modalities and Resolution
• MOSAIC: Generating Consistent, Privacy-Preserving Scenes from Multiple Depth Views in Multi-Room Environments
• EvidMTL: Evidential Multi-Task Learning for Uncertainty-Aware Semantic Surface Mapping from Monocular RGB Images
• Towards Open-Vocabulary Semantic Segmentation Without Semantic Labels
• Privacy-Aware Visual Language Models
• Breaking the Layer Barrier: Remodeling Private Transformer Inference with Hybrid CKKS and MPC
• EMCAD: Efficient Multi-scale Convolutional Attention Decoding for Medical Image Segmentation
• Improving Bird's Eye View Semantic Segmentation by Task Decomposition
• AM-RADIO: Agglomerative Vision Foundation Model -- Reduce All Domains Into One
• VLFM: Vision-Language Frontier Maps for Zero-Shot Semantic Navigation
• Using super-resolution for enhancing visual perception and segmentation performance in veterinary cytology
• DFormer: Diffusion-guided Transformer for Universal Image Segmentation
• Medical SAM Adapter: Adapting Segment Anything Model for Medical Image Segmentation
• Micrograph segmentations for DDEVD
• Segment Anything
• Semantically Accurate Super-Resolution Generative Adversarial Networks
• Masked-attention Mask Transformer for Universal Image Segmentation
• Anonymization for Skeleton Action Recognition
• SegFormer: Simple and Efficient Design for Semantic Segmentation with\n Transformers
• Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers
• Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers
• You Only Need Adversarial Supervision for Semantic Image Synthesis
• ENSEI: Efficient Secure Inference via Frequency-Domain Homomorphic Convolution for Privacy-Preserving Visual Recognition
• An End-to-end Framework For Low-Resolution Remote Sensing Semantic Segmentation
• DeepPrivacy: A Generative Adversarial Network for Face Anonymization
• Deep High-Resolution Representation Learning for Visual Recognition
• Deep High-Resolution Representation Learning for Visual Recognition
• ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
• Spectral Normalization for Generative Adversarial Networks
• Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
• Connecting Pixels to Privacy and Utility: Automatic Redaction of Private\n Information in Images
• The Perception-Distortion Tradeoff
• Extreme Low Resolution Activity Recognition with Multi-Siamese Embedding\n Learning
• Pyramid Scene Parsing Network
• Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
• Fully Convolutional Networks for Semantic Segmentation
• Very Deep Convolutional Networks for Large-Scale Image Recognition

Related

• Semantic Segmentation from Limited Training Data
• Semantic RGB-D Image Synthesis
• SEMNAV: A Semantic Segmentation-Driven Approach to Visual Semantic Navigation
• Semantic Segmentation and Scene Reconstruction of RGB-D Image Frames: An End-to-End Modular Pipeline for Robotic Applications
• Find my mug: Efficient object search with a mobile robot using semantic segmentation