# Privacy Enhanced Multimodal Neural Representations for Emotion Recognition

arXiv: https://arxiv.org/abs/1910.13212 · HTML: https://mlsanity.com/paper/1910.13212/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Oct 2019.

## Built on by (most cited first, 22 of 22)

- [MMANet: Margin-aware Distillation and Modality-aware Regularization for Incomplete Multimodal Learning](https://mlsanity.com/paper/2304.08028/) (2023)
- [Disparate Impact in Differential Privacy from Gradient Misalignment](https://mlsanity.com/paper/2206.07737/) (2022)
- [Fairness and Bias in Multimodal AI: A Survey](https://mlsanity.com/paper/2406.19097/) (2024)
- [Does CLIP Know My Face?](https://mlsanity.com/paper/2209.07341/) (2022)
- [Membership Inference Attacks on Large-Scale Models: A Survey](https://mlsanity.com/paper/2503.19338/) (2025)
- [A Review of Speech-centric Trustworthy Machine Learning: Privacy, Safety, and Fairness](https://mlsanity.com/paper/2212.09006/) (2022)
- [FRAMU: Attention-based Machine Unlearning using Federated Reinforcement Learning](https://mlsanity.com/paper/2309.10283/) (2023)
- [Attribute Inference Attack of Speech Emotion Recognition in Federated Learning Settings](https://mlsanity.com/paper/2112.13416/) (2021)
- [You Are What You Write: Preserving Privacy in the Era of Large Language Models](https://mlsanity.com/paper/2204.09391/) (2022)
- [Privacy-Preserving Video Classification with Convolutional Neural Networks](https://mlsanity.com/paper/2102.03513/) (2021)
- [Configurable Privacy-Preserving Automatic Speech Recognition](https://mlsanity.com/paper/2104.00766/) (2021)
- [MCMoE: Completing Missing Modalities with Mixture of Experts for Incomplete Multimodal Action Quality Assessment](https://mlsanity.com/paper/2511.17397/) (2025)
- [One-stage Modality Distillation for Incomplete Multimodal Learning](https://mlsanity.com/paper/2309.08204/) (2023)
- [Fairness-aware Summarization for Justified Decision-Making](https://mlsanity.com/paper/2107.06243/) (2021)
- [Safeguarding Privacy in Edge Speech Understanding with Tiny Foundation Models](https://mlsanity.com/paper/2502.01649/) (2025)
- [An Attribute-Aligned Strategy for Learning Speech Representation](https://mlsanity.com/paper/2106.02810/) (2021)
- [Don't Judge Me by My Face : An Indirect Adversarial Approach to Remove Sensitive Information From Multimodal Neural Representation in Asynchronous Job Video Interviews](https://mlsanity.com/paper/2110.09424/) (2021)
- [ICYM2I: The illusion of multimodal informativeness under missingness](https://mlsanity.com/paper/2505.16953/) (2025)
- [Human-Imitating Metrics for Training and Evaluating Privacy Preserving Emotion Recognition Models Using Sociolinguistic Knowledge](https://mlsanity.com/paper/2104.08792/) (2021)
- [Private Training Set Inspection in MLaaS](https://mlsanity.com/paper/2305.09058/) (2023)
- [Partial Federated Learning](https://mlsanity.com/paper/2403.01615/) (2024)
- [Reliability-aware Cross-sample Enhancement for Robust Multimodal Sentiment Analysis](https://mlsanity.com/paper/2609.30470/) (2026)

## Abstract (arXiv)

Many mobile applications and virtual conversational agents now aim to recognize and adapt to emotions. To enable this, data are transmitted from users' devices and stored on central servers. Yet, these data contain sensitive information that could be used by mobile applications without user's consent or, maliciously, by an eavesdropping adversary. In this work, we show how multimodal representations trained for a primary task, here emotion recognition, can unintentionally leak demographic information, which could override a selected opt-out option by the user. We analyze how this leakage differs in representations obtained from textual, acoustic, and multimodal data. We use an adversarial learning paradigm to unlearn the private information present in a representation and investigate the effect of varying the strength of the adversarial component on the primary task and on the privacy metric, defined here as the inability of an attacker to predict specific demographic information. We evaluate this paradigm on multiple datasets and show that we can improve the privacy metric while not significantly impacting the performance on the primary task. To the best of our knowledge, this is the first work to analyze how the privacy metric differs across modalities and how multiple privacy concerns can be tackled while still maintaining performance on emotion recognition.