# On Single Source Robustness in Deep Fusion Models

arXiv: https://arxiv.org/abs/1906.04691 · HTML: https://mlsanity.com/paper/1906.04691/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Jun 2019.

## Built on by (most cited first, 12 of 12)

- [One Perturbation is Enough: On Generating Universal Adversarial Perturbations against Vision-Language Pre-training Models](https://mlsanity.com/paper/2406.05491/) (2024)
- [Self-supervised Learning from 100 Million Medical Images](https://mlsanity.com/paper/2201.01283/) (2022)
- [Defending Multimodal Fusion Models against Single-Source Adversaries](https://mlsanity.com/paper/2206.12714/) (2022)
- [Fusion is Not Enough: Single Modal Attacks on Fusion Models for 3D Object Detection](https://mlsanity.com/paper/2304.14614/) (2023)
- [Sensor Adversarial Traits: Analyzing Robustness of 3D Object Detection Sensor Fusion Models](https://mlsanity.com/paper/2109.06363/) (2021)
- [What Makes for Robust Multi-Modal Models in the Face of Missing Modalities?](https://mlsanity.com/paper/2310.06383/) (2023)
- [MMCert: Provable Defense against Adversarial Attacks to Multi-modal Models](https://mlsanity.com/paper/2403.19080/) (2024)
- [Efficient and Effective Universal Adversarial Attack against Vision-Language Pre-training Models](https://mlsanity.com/paper/2410.11639/) (2024)
- [Training-Free Robust Multimodal Learning via Sample-Wise Jacobian Regularization](https://mlsanity.com/paper/2204.02485/) (2022)
- [Informative Data Selection with Uncertainty for Multi-modal Object Detection](https://mlsanity.com/paper/2304.11697/) (2023)
- [Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning](https://mlsanity.com/paper/2305.11347/) (2023)
- [LAMP: Learning Universal Adversarial Perturbations for Multi-Image Tasks via Pre-trained Models](https://mlsanity.com/paper/2601.21220/) (2026)

## Abstract (arXiv)

Algorithms that fuse multiple input sources benefit from both complementary and shared information. Shared information may provide robustness against faulty or noisy inputs, which is indispensable for safety-critical applications like self-driving cars. We investigate learning fusion algorithms that are robust against noise added to a single source. We first demonstrate that robustness against single source noise is not guaranteed in a linear fusion model. Motivated by this discovery, two possible approaches are proposed to increase robustness: a carefully designed loss with corresponding training algorithms for deep fusion models, and a simple convolutional fusion layer that has a structural advantage in dealing with noise. Experimental results show that both training algorithms and our fusion layer make a deep fusion-based 3D object detector robust against noise applied to a single source, while preserving the original performance on clean data.