mlsanityData to 14 Oct 2026

FIDA: Feature Instability-Driven Attack on Self-Supervised Facial Representation

arXiv 2608.26861 ↗First posted 27 Aug 2026cs.CV cs.CRcited by 0

On the map: LLM safety and security → Backdoor and data poisoning attacks

Abstract (arXiv)

Self-supervised learning (SSL) models are vulnerable to backdoor attacks. However, the systemic risks they pose in face representation have received little attention. The entanglement of identity features in self-supervised face learning presents unique challenges for attack stealthiness. To address this gap, we propose FIDA (Feature Instability-Driven Attack), a novel backdoor attack framework. FIDA uses subtle semantic triggers for injection, but its key innovation is a novel objective called Feature Instability Loss. It trains the encoder to increase the sensitivity of triggered features along perturbation directions sampled during attack optimization . By preventing the backdoor from exhibiting the rigid feature patterns typical of previous attacks, FIDA effectively evades the evaluated perturbation-based defenses. Experiments show that FIDA achieves a high attack success rate and generally preserves benign utility across the evaluated settings , posing a significant threat to real-world multimedia applications relying on facial analysis.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security. Builds on 32 of its 49 references that are in the map. First posted on arXiv 27 Aug 2026.

Builds on

32 of 49 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025BadVideo: Stealthy Backdoor Attack against Text-to-Video Generation23 Apr
  2. 2025Secure Transfer Learning: Training Clean Models Against Backdoor in (Both) Pre-trained Encoders and Downstream Datasets16 Apr
  3. 2024Mutual Information Guided Backdoor Mitigation for Pre-trained Encoders5 Jun
  4. 2023Text-to-Image Diffusion Models can be Easily Backdoored through Multimodal Data Poisoning7 May
  5. 2023Detecting Backdoors in Pre-trained Encoders23 Mar
  6. 2022CorruptEncoder: Data Poisoning based Backdoor Attacks to Contrastive Learning15 Nov

Other references · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  3. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  4. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  5. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  6. 2020Exploring Simple Siamese Representation Learningcited by 2,866
  7. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
  8. 2021Barlow Twins: Self-Supervised Learning via Redundancy Reductioncited by 1,838
  9. 2017VGGFace2: A dataset for recognising faces across pose and agecited by 914
  10. 2019STRIP: A Defence Against Trojan Attacks on Deep Neural Networkscited by 459
  11. 2017Age Progression/Regression by Conditional Adversarial Autoencodercited by 429
  12. 2021Backdoor Attacks on Self-Supervised Learningcited by 90
14 more
  1. 2019Bypassing Backdoor Detection Algorithms in Deep Learningcited by 74
  2. 2022The "Beatrix'' Resurrections: Robust Backdoor Detection via Gram Matricescited by 63
  3. 2024AVFF: Audio-Visual Feature Fusion for Video Deepfake Detectioncited by 59
  4. 2024Generalizing Deepfake Video Detection with Plug-and-Play: Video-Level Blending and Spatiotemporal Adapter Tuningcited by 59
  5. 2023ASSET: Robust Backdoor Data Detection Across a Multiplicity of Deep Learning Paradigmscited by 29
  6. 2023Robust Backdoor Detection for Deep Learning via Topological Evolution Dynamicscited by 26
  7. 2022Pose-disentangled Contrastive Learning for Self-supervised Facial Representationcited by 17
  8. 2024Self-Supervised Facial Representation Learning with Facial Region Awarenesscited by 17
  9. 2024Backdoor Contrastive Learning via Bi-level Trigger Optimizationcited by 14
  10. 2024LOTUS: Evasive and Resilient Backdoor Attacks through Sub-Partitioningcited by 13
  11. 2025Backdooring Self-Supervised Contrastive Learning by Noisy Alignmentcited by 6
  12. 2025Rethinking Individual Fairness in Deepfake Detectioncited by 4
  13. 2025PE-CLIP: A Parameter-Efficient Fine-Tuning of Vision Language Models for Dynamic Facial Expression Recognitioncited by 3
  14. 2023Contrastive Learning of View-Invariant Representations for Facial Expressions Recognitioncited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Rethinking the Vulnerabilities of Face Recognition Systems:From a Practical Perspective
  2. 2024FSFM: A Generalizable Face Security Foundation Model via Self-Supervised Facial Representation Learning
  3. 2024MakeupAttack: Feature Space Black-box Backdoor Attack on Face Recognition via Makeup Transfer
  4. 2025SoK: On the Survivability of Backdoor Attacks on Unconstrained Face Recognition Systems
  5. 2025Scalable Face Security Vision Foundation Model for Deepfake, Diffusion, and Spoofing Detection
  6. 2022Resurrecting Trust in Facial Recognition: Mitigating Backdoor Attacks in Face Recognition to Prevent Potential Privacy Breaches
  7. 2024Invisible Backdoor Attack against Self-supervised Learning
  8. 2024EmInspector: Combating Backdoor Attacks in Federated Self-Supervised Learning Through Embedding Inspection