# Cross-Attention is Not Enough: Incongruity-Aware Dynamic Hierarchical Fusion for Multimodal Affect Recognition

arXiv: https://arxiv.org/abs/2305.13583 · HTML: https://mlsanity.com/paper/2305.13583/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 47 references that are in the map.
- First posted on arXiv 23 May 2023.

## Builds on (most cited first, 28 of 28)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks](https://mlsanity.com/paper/1908.02265/) (2019)
- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [Multimodal Machine Learning: A Survey and Taxonomy](https://mlsanity.com/paper/1705.09406/) (2017)
- [Multimodal Transformer for Unaligned Multimodal Language Sequences](https://mlsanity.com/paper/1906.00295/) (2019)
- [Tensor Fusion Network for Multimodal Sentiment Analysis](https://mlsanity.com/paper/1707.07250/) (2017)
- [Efficient Low-rank Multimodal Fusion with Modality-Specific Factors](https://mlsanity.com/paper/1806.00064/) (2018)
- [MISA: Modality-Invariant and -Specific Representations for Multimodal Sentiment Analysis](https://mlsanity.com/paper/2005.03545/) (2020)
- [Towards Multimodal Sarcasm Detection (An _Obviously_ Perfect Paper)](https://mlsanity.com/paper/1906.01815/) (2019)
- [Learning Factorized Multimodal Representations](https://mlsanity.com/paper/1806.06176/) (2018)
- [Memory Fusion Network for Multi-view Sequential Learning](https://mlsanity.com/paper/1802.00927/) (2018)
- [What Makes Multi-modal Learning Better than Single (Provably)](https://mlsanity.com/paper/2106.04538/) (2021)
- [Found in Translation: Learning Robust Joint Representations by Cyclic Translations Between Modalities](https://mlsanity.com/paper/1812.07809/) (2018)
- [Multi-attention Recurrent Network for Human Communication Comprehension](https://mlsanity.com/paper/1802.00923/) (2018)
- [Words Can Shift: Dynamically Adjusting Word Representations Using Nonverbal Behaviors](https://mlsanity.com/paper/1811.09362/) (2018)
- [Multimodal Language Analysis with Recurrent Multistage Fusion](https://mlsanity.com/paper/1808.03920/) (2018)
- [Multimodal Learning using Optimal Transport for Sarcasm and Humor Detection](https://mlsanity.com/paper/2110.10949/) (2021)
- [Gold Doesn't Always Glitter: Spectral Removal of Linear and Nonlinear Guarded Attribute Information](https://mlsanity.com/paper/2203.07893/) (2022)
- [Nice perfume. How long did you marinate in it? Multimodal Sarcasm Explanation](https://mlsanity.com/paper/2112.04873/) (2021)
- [CARCA: Context and Attribute-Aware Next-Item Recommendation via Cross-Attention](https://mlsanity.com/paper/2204.06519/) (2022)
- [Fusing ASR Outputs in Joint Training for Speech Emotion Recognition](https://mlsanity.com/paper/2110.15684/) (2021)
- [Factorized Multimodal Transformer for Multimodal Sequential Learning](https://mlsanity.com/paper/1911.09826/) (2019)
- [Is Cross-Attention Preferable to Self-Attention for Multi-Modal Emotion Recognition?](https://mlsanity.com/paper/2202.09263/) (2022)
- [Low Rank Fusion based Transformers for Multimodal Sequences](https://mlsanity.com/paper/2007.02038/) (2020)
- [On the Use of Modality-Specific Large-Scale Pre-Trained Encoders for Multimodal Sentiment Analysis](https://mlsanity.com/paper/2210.15937/) (2022)
- [Cross-Attention is all you need: Real-Time Streaming Transformers for Personalised Speech Enhancement](https://mlsanity.com/paper/2211.04346/) (2022)
- [Multimodal Dyadic Impression Recognition via Listener Adaptive Cross-Domain Fusion](https://mlsanity.com/paper/2211.05163/) (2022)

## Built on by (most cited first, 1 of 1)

- [Asynchronous Multimodal Video Sequence Fusion via Learning Modality-Exclusive and -Agnostic Representations](https://mlsanity.com/paper/2407.04955/) (2024)

## Abstract (arXiv)

Fusing multiple modalities has proven effective for multimodal information processing. However, the incongruity between modalities poses a challenge for multimodal fusion, especially in affect recognition. In this study, we first analyze how the salient affective information in one modality can be affected by the other, and demonstrate that inter-modal incongruity exists latently in crossmodal attention. Based on this finding, we propose the Hierarchical Crossmodal Transformer with Dynamic Modality Gating (HCT-DMG), a lightweight incongruity-aware model, which dynamically chooses the primary modality in each training batch and reduces fusion times by leveraging the learned hierarchy in the latent space to alleviate incongruity. The experimental evaluation on five benchmark datasets: CMU-MOSI, CMU-MOSEI, and IEMOCAP (sentiment and emotion), where incongruity implicitly lies in hard samples, as well as UR-FUNNY (humour) and MUStaRD (sarcasm), where incongruity is common, verifies the efficacy of our approach, showing that HCT-DMG: 1) outperforms previous multimodal models with a reduced size of approximately 0.8M parameters; 2) recognizes hard samples where incongruity makes affect recognition difficult; 3) mitigates the incongruity at the latent level in crossmodal attention.