# Multimodal Ambivalence and Hesitancy Recognition via Cross-Attention and Gated Fusion

arXiv: https://arxiv.org/abs/2607.15779 · HTML: https://mlsanity.com/paper/2607.15779/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 21 references that are in the map.
- First posted on arXiv 17 Jul 2026.

## Builds on (most cited first, 20 of 20)

- [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)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Optuna: A Next-generation Hyperparameter Optimization Framework](https://mlsanity.com/paper/1907.10902/) (2019)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [Video-LLaVA: Learning United Visual Representation by Alignment Before Projection](https://mlsanity.com/paper/2311.10122/) (2023)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking](https://mlsanity.com/paper/2303.16727/) (2023)
- [BAH Dataset for Ambivalence/Hesitancy Recognition in Videos for Digital Behavioural Change](https://mlsanity.com/paper/2505.19328/) (2025)
- [F2LLM-v2: Inclusive, Performant, and Efficient Embeddings for a Multilingual World](https://mlsanity.com/paper/2603.19223/) (2026)
- [From Affect to Complex Behavior: Advancing Multimodal Human-Centered AI at the 10th ABAW Workshop & Competition](https://mlsanity.com/paper/2605.27451/) (2026)
- [Conflict-Aware Multimodal Fusion for Ambivalence and Hesitancy Recognition](https://mlsanity.com/paper/2603.15818/) (2026)
- [F2LLM Technical Report: Matching SOTA Embedding Performance with 6 Million Open-Source Data](https://mlsanity.com/paper/2510.02294/) (2025)
- [BROTHER: Behavioral Recognition Optimized Through Heterogeneous Ensemble Regularization for Ambivalence and Hesitancy](https://mlsanity.com/paper/2603.14361/) (2026)
- [Team LEYA in 10th ABAW Competition: Multimodal Ambivalence/Hesitancy Recognition Approach](https://mlsanity.com/paper/2603.12848/) (2026)
- [Solution for 10th Competition on Ambivalence/Hesitancy (AH) Video Recognition Challenge using Divergence-Based Multimodal Fusion](https://mlsanity.com/paper/2603.16939/) (2026)
- [Multimodal Ambivalence/Hesitancy Recognition in Videos for Personalized Digital Health Interventions](https://mlsanity.com/paper/2604.11730/) (2026)

## Abstract (arXiv)

We present a multimodal framework for Ambivalence/Hesitancy (A/H) recognition in video, developed for the ABAW11 challenge at ECCV 2026. The proposed approach fuses textual, acoustic, and visual modalities extracted from the BAH dataset using three pretrained encoders: F2LLM-v2-0.6B for transcripts (1024-d), WavLM-Large for audio (1024-d), and VideoMAE V2 for facial video (768-d). We first establish comprehensive unimodal baselines using classical classifiers (MLP, Random Forest, GBDT), each optimized via Optuna, and obtain a best unimodal Macro F1 of 0.6659 on the test set using text features alone -- substantially outperforming the zero-shot Video-LLaVA baseline (Macro F1: 0.2827). Building on these baselines, we propose a multimodal fusion architecture that combines bidirectional cross-attention across all three modalities with a Gated Multimodal Unit (GMU), with both architectural and optimization hyperparameters selected through a 50-trial Optuna search. This model achieves a Macro F1 of 0.7394 on the validation set, a relative improvement of 11.0% over the best unimodal baseline, confirming that explicit cross-modal interaction captures complementary cues that no single modality provides in isolation. Final predictions on the official, unlabeled private test set are generated using this model and submitted according to the challenge protocol. Code is publicly available at https://github.com/yassineouzar/IUSDAH/