CleanDIFT: Diffusion Features without Noise
This work addresses a bottleneck for researchers and practitioners using diffusion features in downstream tasks, offering a more efficient and effective solution, though it is incremental as it builds on existing diffusion model frameworks.
The paper tackles the problem that diffusion models require adding noise to images to extract useful semantic features, which harms performance and cannot be fixed by ensembling. It introduces a lightweight fine-tuning method to produce noise-free features, achieving better performance than previous methods in various tasks, such as outperforming ensemble-based approaches at lower cost.
Internal features from large-scale pre-trained diffusion models have recently been established as powerful semantic descriptors for a wide range of downstream tasks. Works that use these features generally need to add noise to images before passing them through the model to obtain the semantic features, as the models do not offer the most useful features when given images with little to no noise. We show that this noise has a critical impact on the usefulness of these features that cannot be remedied by ensembling with different random noises. We address this issue by introducing a lightweight, unsupervised fine-tuning method that enables diffusion backbones to provide high-quality, noise-free semantic features. We show that these features readily outperform previous diffusion features by a wide margin in a wide variety of extraction setups and downstream tasks, offering better performance than even ensemble-based methods at a fraction of the cost.