MultiDelete for Multimodal Machine Unlearning
It addresses the challenge of efficiently removing specific knowledge from trained multimodal models, which is crucial for privacy and data management in AI applications, representing a novel approach in this domain.
The paper tackles the problem of machine unlearning in multimodal settings by introducing MultiDelete, which decouples associations between unimodal data points during unlearning, resulting in an average improvement of 17.6 points over baselines in unlearning multimodal samples while maintaining model knowledge and providing better protection against attacks.
Machine Unlearning removes specific knowledge about training data samples from an already trained model. It has significant practical benefits, such as purging private, inaccurate, or outdated information from trained models without the need for complete re-training. Unlearning within a multimodal setting presents unique challenges due to the complex dependencies between different data modalities and the expensive cost of training on large multimodal datasets and architectures. This paper presents the first machine unlearning approach for multimodal data and models, titled MultiDelete, which is designed to decouple associations between unimodal data points during unlearning without losing the overall representation strength of the trained model. MultiDelete advocates for three key properties for effective multimodal unlearning: (a): modality decoupling, which effectively decouples the association between individual unimodal data points marked for deletion, rendering them as unrelated data points, (b): multimodal knowledge retention, which retains the multimodal representation post-unlearning, and (c): unimodal knowledge retention, which retains the unimodal representation postunlearning. MultiDelete is efficient to train and is not constrained by using a strongly convex loss -- a common restriction among existing baselines. Experiments on two architectures and four datasets, including image-text and graph-text datasets, show that MultiDelete gains an average improvement of 17.6 points over best performing baseline in unlearning multimodal samples, can maintain the multimodal and unimodal knowledge of the original model post unlearning, and can provide better protection to unlearned data against adversarial attacks.