# AI Safety in Practice: Enhancing Adversarial Robustness in Multimodal Image Captioning

arXiv: https://arxiv.org/abs/2407.21174 · HTML: https://mlsanity.com/paper/2407.21174/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 20 references that are in the map.
- First posted on arXiv 30 Jul 2024.

## Builds on (most cited first, 9 of 9)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [Towards Evaluating the Robustness of Neural Networks](https://mlsanity.com/paper/1608.04644/) (2016)
- [The Limitations of Deep Learning in Adversarial Settings](https://mlsanity.com/paper/1511.07528/) (2015)
- [Adversarial Training for Free!](https://mlsanity.com/paper/1904.12843/) (2019)
- [On the Adversarial Robustness of Multi-Modal Foundation Models](https://mlsanity.com/paper/2308.10741/) (2023)
- [Reading Isn't Believing: Adversarial Attacks On Multi-Modal Neurons](https://mlsanity.com/paper/2103.10480/) (2021)
- [Investigating Vulnerability to Adversarial Examples on Multimodal Data Fusion in Deep Learning](https://mlsanity.com/paper/2005.10987/) (2020)
- [Adversarial Evaluation of Multimodal Models under Realistic Gray Box Assumption](https://mlsanity.com/paper/2011.12902/) (2020)

## Built on by (most cited first, 3 of 3)

- [Dual-Modality Multi-Stage Adversarial Safety Training: Robustifying Multimodal Web Agents Against Cross-Modal Attacks](https://mlsanity.com/paper/2603.04364/) (2026)
- [A Framework for Evaluating Vision-Language Model Safety: Building Trust in AI for Public Sector Applications](https://mlsanity.com/paper/2502.16361/) (2025)
- [Exploring Visual Embedding Spaces Induced by Vision Transformers for Online Auto Parts Marketplaces](https://mlsanity.com/paper/2502.05756/) (2025)

## Abstract (arXiv)

Multimodal machine learning models that combine visual and textual data are increasingly being deployed in critical applications, raising significant safety and security concerns due to their vulnerability to adversarial attacks. This paper presents an effective strategy to enhance the robustness of multimodal image captioning models against such attacks. By leveraging the Fast Gradient Sign Method (FGSM) to generate adversarial examples and incorporating adversarial training techniques, we demonstrate improved model robustness on two benchmark datasets: Flickr8k and COCO. Our findings indicate that selectively training only the text decoder of the multimodal architecture shows performance comparable to full adversarial training while offering increased computational efficiency. This targeted approach suggests a balance between robustness and training costs, facilitating the ethical deployment of multimodal AI systems across various domains.