SPARE: Self-distillation for PARameter-Efficient Removal
Machine Unlearning aims to remove the influence of specific data or concepts from trained models while preserving overall performance, a capability increasingly required by data protection regulations and responsible AI practices. Despite recent progress, unlearning in text-to-image diffusion models remains challenging due to high computational costs and the difficulty of balancing effective forgetting with retention of unrelated concepts. We introduce Self-distillation for PARameter Efficient Removal (SPARE), a two-stage unlearning method for image generation that combines parameter localization with self-distillation. SPARE first identifies parameters most responsible for generation of the unwanted concepts using gradient-based saliency and constrains updates through sparse low rank adapters, ensuring lightweight, localized modifications. In a second stage, SPARE applies a self-distillation objective that overwrites the unwanted concept with a user-defined surrogate while preserving behavior for other concepts. In addition we proposed a timestep sampling scheme for diffusion models to target only the crucial timesteps for a given concept leading to efficient unlearning. SPARE surpasses the current state-of-the-art on the UnlearnCanvas benchmark, and ablation studies on several datasets indicate fine-grained control over the forgetting-retention trade-off. Our results demonstrate that SPARE achieves strong concept erasure and high retainability across various domains, making it a suitable solution for selective unlearning in diffusion-based image generation models.
Code (0)
등록된 구현이 없습니다.
Tasks
Image GenerationSimilar Papers 제목 키워드 기반
JSTASR: Joint Size and Transparency-Aware Snow Removal Algorithm Based on Modified Partial Convolution and Veiling Effect Removal
Snow removal usually affects the performance of computer vision. Comparing with other atmospheric phenomenon (e.g., haze and rain), snow is more complicated due to its transparency, various size, and accumulation of veil…
object-detectionObject DetectionSnow RemovalRethinking Feature-Based Knowledge Distillation for Face Recognition
With the continual expansion of face datasets, feature-based distillation prevails for large-scale face recognition. In this work, we attempt to remove identity supervision in student training, to spare the GPU memor…
Face RecognitionGPUKnowledge DistillationCan LLM Watermarks Robustly Prevent Unauthorized Knowledge Distillation?
The radioactive nature of Large Language Model (LLM) watermarking enables the detection of watermarks inherited by student models when trained on the outputs of watermarked teacher models, making it a promising tool for …
Knowledge DistillationLanguage ModelingLanguage ModellingLarge Language Model+1Single-image reflection removal via self-supervised diffusion models
Reflections often degrade the visual quality of images captured through transparent surfaces, and reflection removal methods suffers from the shortage of paired real-world samples.This paper proposes a hybrid approach th…
DenoisingReflection RemovalContinuous Adverse Weather Removal via Degradation-Aware Distillation
All-in-one models for adverse weather removal aim to process various degraded images using a single set of parameters, making them ideal for real-world scenarios. However, they encounter two main challenges: catastro…
Incremental Learning