This paper provides a comprehensive survey of multimodal fusion techniques, focusing on the challenges and recent advancements in handling low-quality data. The authors identify four main challenges in multimodal fusion on low-quality data: noisy multimodal data, incomplete multimodal data, imbalanced multimodal data, and quality-varying multimodal data. They present a new taxonomy to help researchers understand the state of the field and explore potential research directions. The paper discusses various approaches to address these challenges, including noise reduction, imputation methods, and techniques for handling imbalanced and varying quality data. The authors also highlight the importance of leveraging cross-modal correlations and complementarity to improve the robustness and reliability of multimodal fusion in real-world applications.This paper provides a comprehensive survey of multimodal fusion techniques, focusing on the challenges and recent advancements in handling low-quality data. The authors identify four main challenges in multimodal fusion on low-quality data: noisy multimodal data, incomplete multimodal data, imbalanced multimodal data, and quality-varying multimodal data. They present a new taxonomy to help researchers understand the state of the field and explore potential research directions. The paper discusses various approaches to address these challenges, including noise reduction, imputation methods, and techniques for handling imbalanced and varying quality data. The authors also highlight the importance of leveraging cross-modal correlations and complementarity to improve the robustness and reliability of multimodal fusion in real-world applications.