Dreamguider: Improved Training free Diffusion-based Conditional Generation
This work addresses a key bottleneck in training-free diffusion models for researchers and practitioners, offering a lightweight solution for both linear and non-linear guidance problems, though it is incremental in improving existing inference-time techniques.
The paper tackles the problem of compute-heavy backpropagation and handcrafted parameter tuning in diffusion-based conditional generation by proposing Dreamguider, a method that uses a time-varying factor to regulate gradient flow and an empirical guidance scale, achieving effective performance across multiple tasks and datasets.
Diffusion models have emerged as a formidable tool for training-free conditional generation.However, a key hurdle in inference-time guidance techniques is the need for compute-heavy backpropagation through the diffusion network for estimating the guidance direction. Moreover, these techniques often require handcrafted parameter tuning on a case-by-case basis. Although some recent works have introduced minimal compute methods for linear inverse problems, a generic lightweight guidance solution to both linear and non-linear guidance problems is still missing. To this end, we propose Dreamguider, a method that enables inference-time guidance without compute-heavy backpropagation through the diffusion network. The key idea is to regulate the gradient flow through a time-varying factor. Moreover, we propose an empirical guidance scale that works for a wide variety of tasks, hence removing the need for handcrafted parameter tuning. We further introduce an effective lightweight augmentation strategy that significantly boosts the performance during inference-time guidance. We present experiments using Dreamguider on multiple tasks across multiple datasets and models to show the effectiveness of the proposed modules. To facilitate further research, we will make the code public after the review process.