# VAEmo: Efficient Representation Learning for Visual-Audio Emotion with Knowledge Injection

arXiv: https://arxiv.org/abs/2505.02331 · HTML: https://mlsanity.com/paper/2505.02331/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 56 references that are in the map.
- First posted on arXiv 5 May 2025.

## Builds on (most cited first, 30 of 33)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [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)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [VGGFace2: A dataset for recognising faces across pose and age](https://mlsanity.com/paper/1710.08092/) (2017)
- [VoxCeleb2: Deep Speaker Recognition](https://mlsanity.com/paper/1806.05622/) (2018)
- [NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models](https://mlsanity.com/paper/2405.17428/) (2024)
- [MERLOT: Multimodal Neural Script Knowledge Models](https://mlsanity.com/paper/2106.02636/) (2021)
- [Masked Autoencoders that Listen](https://mlsanity.com/paper/2207.06405/) (2022)
- [Deep Facial Expression Recognition: A Survey](https://mlsanity.com/paper/1804.08348/) (2018)
- [Contrastive Audio-Visual Masked Autoencoder](https://mlsanity.com/paper/2210.07839/) (2022)
- [Meta-Transformer: A Unified Framework for Multimodal Learning](https://mlsanity.com/paper/2307.10802/) (2023)
- [DFEW: A Large-Scale Database for Recognizing Dynamic Facial Expressions in the Wild](https://mlsanity.com/paper/2008.05924/) (2020)
- [Emotion-LLaMA: Multimodal Emotion Recognition and Reasoning with Instruction Tuning](https://mlsanity.com/paper/2406.11161/) (2024)
- [Aff-Wild2: Extending the Aff-Wild Database for Affect Recognition](https://mlsanity.com/paper/1811.07770/) (2018)
- [FERV39k: A Large-Scale Multi-Scene Dataset for Facial Expression Recognition in Videos](https://mlsanity.com/paper/2203.09463/) (2022)
- [MAFW: A Large-scale, Multi-modal, Compound Affective Database for Dynamic Facial Expression Recognition in the Wild](https://mlsanity.com/paper/2208.00847/) (2022)
- [Tuning LayerNorm in Attention: Towards Efficient Multi-Modal LLM Finetuning](https://mlsanity.com/paper/2312.11420/) (2023)
- [MAViL: Masked Audio-Video Learners](https://mlsanity.com/paper/2212.08071/) (2022)
- [MAE-DFER: Efficient Masked Autoencoder for Self-supervised Dynamic Facial Expression Recognition](https://mlsanity.com/paper/2307.02227/) (2023)
- [Intensity-Aware Loss for Dynamic Facial Expression Recognition in the Wild](https://mlsanity.com/paper/2208.10335/) (2022)
- [From Static to Dynamic: Adapting Landmark-Aware Image Models for Facial Expression Recognition in Videos](https://mlsanity.com/paper/2312.05447/) (2023)
- [GPT-4V with Emotion: A Zero-shot Benchmark for Generalized Emotion Recognition](https://mlsanity.com/paper/2312.04293/) (2023)
- [SECap: Speech Emotion Captioning with Large Language Model](https://mlsanity.com/paper/2312.10381/) (2023)
- [HiCMAE: Hierarchical Contrastive Masked Autoencoder for Self-Supervised Audio-Visual Emotion Recognition](https://mlsanity.com/paper/2401.05698/) (2024)
- [FineCLIPER: Multi-modal Fine-grained CLIP for Dynamic Facial Expression Recognition with AdaptERs](https://mlsanity.com/paper/2407.02157/) (2024)
- [EMO-LLaMA: Enhancing Facial Emotion Understanding with Instruction Tuning](https://mlsanity.com/paper/2408.11424/) (2024)
- [SVFAP: Self-supervised Video Facial Affect Perceiver](https://mlsanity.com/paper/2401.00416/) (2023)

## Built on by (most cited first, 5 of 5)

- [Not all Blends are Equal: The BLEMORE Dataset of Blended Emotion Expressions with Relative Salience Annotations](https://mlsanity.com/paper/2601.13225/) (2026)
- [Crab+: A Scalable and Unified Audio-Visual Scene Understanding Model with Explicit Cooperation](https://mlsanity.com/paper/2603.04128/) (2026)
- [Bidirectional Learning of Facial Action Units and Expressions via Structured Semantic Mapping across Heterogeneous Datasets](https://mlsanity.com/paper/2604.10541/) (2026)
- [Ordering Matters: Rank-Aware Selective Fusion for Blended Emotion Recognition](https://mlsanity.com/paper/2605.21417/) (2026)
- [Traits Run Deeper: Trait-Specific Asymmetric Fusion for Multimodal Personality Assessment](https://mlsanity.com/paper/2606.11269/) (2026)

## Abstract (arXiv)

Audiovisual emotion recognition (AVER) aims to infer human emotions from nonverbal visual-audio (VA) cues, offering modality-complementary and language-agnostic advantages. However, AVER remains challenging due to the inherent ambiguity of emotional expressions, cross-modal expressive disparities, and the scarcity of reliably annotated data. Recent self-supervised AVER approaches have introduced strong multimodal representations, yet they predominantly rely on modality-specific encoders and coarse content-level alignment, limiting fine-grained emotional semantic modeling. To address these issues, we propose VAEmo, an efficient two-stage framework for emotion-centric joint VA representation learning with external knowledge injection. In Stage~1, a unified and lightweight representation network is pre-trained on large-scale speaker-centric VA corpora via masked reconstruction and contrastive objectives, mitigating the modality gap and learning expressive, complementary representations without emotion labels. In Stage~2, multimodal large language models automatically generate detailed affective descriptions according to our well-designed chain-of-thought prompting for only a small subset of VA samples; these rich textual semantics are then injected by aligning their corresponding embeddings with VA representations through dual-path contrastive learning, further bridging the emotion gap. Extensive experiments on multiple downstream AVER benchmarks show that VAEmo achieves state-of-the-art performance with a compact design, highlighting the benefit of unified cross-modal encoding and emotion-aware semantic guidance for efficient, generalizable VA emotion representations.