# Multimodal Voice Activity Projection for Turn-Taking in Social Robots with Voice-Activity-Related Pretrained Encoders

arXiv: https://arxiv.org/abs/2607.07294 · HTML: https://mlsanity.com/paper/2607.07294/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 42 references that are in the map.
- First posted on arXiv 8 Jul 2026.

## Builds on (most cited first, 19 of 19)

- [Is Someone Speaking? Exploring Long-term Temporal Features for Audio-visual Active Speaker Detection](https://mlsanity.com/paper/2107.06592/) (2021)
- [Voice Activity Projection: Self-supervised Learning of Turn-taking Events](https://mlsanity.com/paper/2205.09812/) (2022)
- [LoRA-Whisper: Parameter-Efficient and Extensible Multilingual ASR](https://mlsanity.com/paper/2406.06619/) (2024)
- [Whisper-Flamingo: Integrating Visual Features into Whisper for Audio-Visual Speech Recognition and Translation](https://mlsanity.com/paper/2406.10082/) (2024)
- [Multilingual Turn-taking Prediction Using Voice Activity Projection](https://mlsanity.com/paper/2403.06487/) (2024)
- [Real-time and Continuous Turn-taking Prediction Using Voice Activity Projection](https://mlsanity.com/paper/2401.04868/) (2024)
- [Applying General Turn-taking Models to Conversational Human-Robot Interaction](https://mlsanity.com/paper/2501.08946/) (2025)
- [Robust Adaptation of Large Multimodal Models for Retrieval Augmented Hateful Meme Detection](https://mlsanity.com/paper/2502.13061/) (2025)
- [Predicting Turn-Taking and Backchannel in Human-Machine Conversations Using Linguistic, Acoustic, and Visual Signals](https://mlsanity.com/paper/2505.12654/) (2025)
- [Yeah, Un, Oh: Continuous and Real-time Backchannel Prediction with Fine-tuning of Voice Activity Projection](https://mlsanity.com/paper/2410.15929/) (2024)
- [Visual Cues Enhance Predictive Turn-Taking for Two-Party Human Interaction](https://mlsanity.com/paper/2505.21043/) (2025)
- [Multimodal Large Language Models with Fusion Low Rank Adaptation for Device Directed Speech Detection](https://mlsanity.com/paper/2406.09617/) (2024)
- [A Noise-Robust Turn-Taking System for Real-World Dialogue Robots: A Field Experiment](https://mlsanity.com/paper/2503.06241/) (2025)
- [Multilingual Dyadic Interaction Corpus NoXi+J: Toward Understanding Asian-European Non-verbal Cultural Characteristics and their Influences on Engagement](https://mlsanity.com/paper/2409.13726/) (2024)
- [Triadic Multi-party Voice Activity Projection for Turn-taking in Spoken Dialogue Systems](https://mlsanity.com/paper/2507.07518/) (2025)
- [When and How to Express Empathy in Human-Robot Interaction Scenarios](https://mlsanity.com/paper/2509.25200/) (2025)
- [Voice Activity Projection Model with Multimodal Encoders](https://mlsanity.com/paper/2506.03980/) (2025)
- [Multimodal Transformer Models for Turn-taking Prediction: Effects on Conversational Dynamics of Human-Agent Interaction during Cooperative Gameplay](https://mlsanity.com/paper/2503.16432/) (2025)
- [M3SD: Multi-modal, Multi-scenario and Multi-language Speaker Diarization Dataset](https://mlsanity.com/paper/2506.14427/) (2025)

## Abstract (arXiv)

Turn-taking prediction is a key requirement for social robots involved in human-human interaction, particularly in mediator settings, where the robot must anticipate conversational dynamics rather than merely react to pauses. This work presents a Multimodal Voice Activity Projection (MM-VAP) framework that extends the original audio-only VAP formulation to synchronized audio-visual inputs while preserving its self-supervised future-projection objective. The proposed approach builds on pretrained audio-visual backbones originally optimized for speech-related tasks and adapts them through Low-Rank Adaptation to the multimodal turn-taking problem. After independent speaker encoding, an inter-speaker attention stage models the relational dynamics required to project future voice activity. In addition, a semantic consistency loss is introduced to regularize the 256-state output space according to higher-level dialogue activity patterns. Experiments on NoXi and NoXi+J showed improvements over the current baselines, particularly for some turn-taking events. Additional evaluation on the Haru EDR corpus further supported the suitability of this direction for mediation-oriented human-robot interaction.