---
format: "aidr-story-markdown/v1"
id: "29aeea54e2a42ca8ca758e01cf95cd58ad94d2a576b3e0e1814f40e3009ce6f8"
canonical_url: "https://aidr.today/29aeea54?lang=en"
title: "Multimodal Ambivalence and Hesitancy Recognition with Bidirectional Audio--Text Cross-Attention"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-21T04:00:00.000Z"
category: "Research"
topics: ["inference"]
source_urls: ["https://arxiv.org/abs/2610.26868"]
summary: "A new audio-text framework uses pre-trained multilingual BERT and wav2vec 2.0 with bidirectional cross-attention blocks to recognize ambivalence and hesitancy in videos, achieving 72.19% accuracy and 77.95% A/H F1-score on the BAH dataset."
---

# Multimodal Ambivalence and Hesitancy Recognition with Bidirectional Audio\-\-Text Cross\-Attention

> [Open the canonical story](<https://aidr.today/29aeea54?lang=en>)

**Published:** 2026-10-21T04:00:00.000Z
**Category:** Research
**Topics:** inference

## Summary

A new audio\-text framework uses pre\-trained multilingual BERT and wav2vec 2\.0 with bidirectional cross\-attention blocks to recognize ambivalence and hesitancy in videos, achieving 72\.19% accuracy and 77\.95% A/H F1\-score on the BAH dataset\.

## Sources

- [Story source](<https://arxiv.org/abs/2610.26868>)