mlsanityData to 7 Oct 2026

Hijacking Vision-and-Language Navigation Agents with Adversarial Environmental Attacks

arXiv 2412.02795 ↗First posted 3 Dec 2024cs.CV cs.ROcited by 5

Not on the map. Why?

Abstract (arXiv)

Assistive embodied agents that can be instructed in natural language to perform tasks in open-world environments have the potential to significantly impact labor tasks like manufacturing or in-home care -- benefiting the lives of those who come to depend on them. In this work, we consider how this benefit might be hijacked by local modifications in the appearance of the agent's operating environment. Specifically, we take the popular Vision-and-Language Navigation (VLN) task as a representative setting and develop a whitebox adversarial attack that optimizes a 3D attack object's appearance to induce desired behaviors in pretrained VLN agents that observe it in the environment. We demonstrate that the proposed attack can cause VLN agents to ignore their instructions and execute alternative actions after encountering the attack object -- even for instructions and agent paths not considered when optimizing the attack. For these novel settings, we find our attacks can induce early-termination behaviors or divert an agent along an attacker-defined multi-step trajectory. Under both conditions, environmental attacks significantly reduce agent capabilities to successfully follow user instructions.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 33 of its 47 references that are in the map. First posted on arXiv 3 Dec 2024.

Builds on

33 of 47 references.

References · 33

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  4. 2013Intriguing properties of neural networkscited by 5,114
  5. 2016Adversarial examples in the physical worldcited by 1,762
  6. 2017Matterport3D: Learning from RGB-D Data in Indoor Environmentscited by 1,354
  7. 2017Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environmentscited by 1,013
  8. 2020Accelerating 3D Deep Learning with PyTorch3Dcited by 781
  9. 2016Universal adversarial perturbationscited by 774
  10. 2017Adversarial Patchcited by 514
  11. 2020Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense Spatiotemporal Groundingcited by 401
  12. 2017Embodied Question Answeringcited by 335
21 more
  1. 2019REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environmentscited by 330
  2. 2017Adversarial Attacks on Neural Network Policiescited by 314
  3. 2023On Evaluating Adversarial Robustness of Large Vision-Language Modelscited by 281
  4. 2021History Aware Multimodal Transformer for Vision-and-Language Navigationcited by 268
  5. 2020Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observationscited by 235
  6. 2017Tactics of Adversarial Attack on Deep Reinforcement Learning Agentscited by 163
  7. 2019Making an Invisibility Cloak: Real World Adversarial Attacks on Object Detectorscited by 155
  8. 2021Dual Attention Suppression Attack: Generate Adversarial Camouflage in Physical Worldcited by 127
  9. 2020Adversarial Camouflage: Hiding Physical-World Attacks with Natural Stylescited by 126
  10. 2022Vision-and-Language Navigation: A Survey of Tasks, Methods, and Future Directionscited by 116
  11. 2019General Evaluation for Instruction Conditioned Navigation using Dynamic Time Warpingcited by 102
  12. 2020Sim-to-Real Transfer for Vision-and-Language Navigationcited by 88
  13. 2021Towards Improving Adversarial Training of NLP Modelscited by 79
  14. 2019Adversarial camera stickers: A physical camera-based attack on deep learning systemscited by 69
  15. 2023How Deep Learning Sees the World: A Survey on Adversarial Attacks & Defensescited by 65
  16. 2024An Image Is Worth 1000 Lies: Adversarial Transferability across Prompts on Vision-Language Modelscited by 40
  17. 20213D Adversarial Attacks Beyond Point Cloudcited by 26
  18. 2022Isometric 3D Adversarial Examples in the Physical Worldcited by 25
  19. 2021Token-Modification Adversarial Attacks for Natural Language Processing: A Surveycited by 12
  20. 2023Behavioral Analysis of Vision-and-Language Navigation Agentscited by 8
  21. 2022Consistent Attack: Universal Adversarial Perturbation on Embodied Vision Navigationcited by 6

Built on by

Cited by year (2026 to 7 Oct)

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

5 papers.

  1. 2025Safety of Embodied Navigation: A Surveycited by 9
  2. 2025Shedding Light on VLN Robustness: A Black-box Framework for Indoor Lighting-based Adversarial Attackcited by 1
  3. 2026From Thinker to Society: Security in Hierarchical Autonomy Evolution of AI Agentscited by 1
  4. 2026NavTrust: Benchmarking Trustworthiness for Embodied Navigationcited by 0
  5. 2026AdvNav: Behavior-Guided Black-Box Adversarial Attacks on Vision-Language Navigationcited by 0