Paper deep dive
EMMU: Efficient Information-Level Multimodal Machine Unlearning with High Model Fidelity
Jie Zhang, Jiahui Hou, Tie Xiao, Yunyi Huang, Xiang-Yang Li
Models: vision-language models (unspecified)
Intelligence
Status: succeeded | Model: google/gemini-3.1-flash-lite-preview | Prompt: intel-v1 | Confidence: 93%
Last extracted: 3/11/2026, 1:16:04 AM
Summary
EMMU is an efficient information-level multimodal machine unlearning framework designed for vision-language systems. It improves unlearning efficiency and model fidelity by identifying and updating only the model parameters highly correlated with specific sensitive information, rather than performing data-level unlearning.
Entities (5)
Relation Signals (3)
EMMU → appliesto → Vision-Language Systems
confidence 95% · we design an efficient multimodal machine unlearning (EMMU) framework to address model fidelity in vision-language systems.
EMMU → performs → Machine Unlearning
confidence 95% · EMMU: Efficient Information-Level Multimodal Machine Unlearning
EMMU → evaluatedon → Visual Question Answering
confidence 90% · Evaluations on vision-language tasks, such as Visual Question Answering
Cypher Suggestions (0)
No Cypher suggestions yet.
Abstract
To comply with the “right to be forgotten,” recent research has introduced machine unlearning techniques that enable machine learning models to remove specific data samples. However, existing multimodal machine unlearning has concerns about efficiency, and the model fidelity may deteriorate after unlearning, leading to meaningless outputs if given data samples that are requested to be forgotten. Instead of focusing on datalevel machine unlearning, we focus on information-level unlearning, aiming to forget specific information of data samples (such as sensitive information involving name or medical condition) while maintaining the model fidelity. In this work, we design an efficient multimodal machine unlearning (EMMU) framework to address model fidelity in vision-language systems. The core idea is to locate and only modify model parameters that are highly correlated with the specific information (which requires forgetting). EMMU locates crucial parameters associated with the sensitive information and updates these parameters using a multiobjective optimization strategy. Evaluations on vision-language tasks, such as Visual Question Answering and Image Captioning, using multiple datasets, demonstrate the efficiency and fidelity of EMMU. Compared to existing methods, our method obtains an average of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$111 \times$</tex> improvement, boosting efficiency up to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1445 \times$</tex>. Meanwhile, the average utility-forget balance score has improved <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$9 \times$</tex> on average and reached up to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$70 \times$</tex>, across multiple models and datasets.
Tags
Links
Full Text
837 characters extracted from source content.
Expand or collapse full text
EMMU: Efficient Information-Level Multimodal Machine Unlearning with High Model Fidelity | IEEE Conference Publication | IEEE Xplore IEEE Account Change Username/Password Update Address Purchase Details Payment Options Order History View Purchased Documents Profile Information Communications Preferences Profession and Education Technical Interests Need Help? US & Canada: +1 800 678 4333 Worldwide: +1 732 981 0060 Contact & Support About IEEE Xplore Contact Us Help Accessibility Terms of Use Nondiscrimination Policy Sitemap Privacy & Opting Out of Cookies A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity.© Copyright 2026 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.