DiffMorpher: Unleashing the Capability of Diffusion Models for Image Morphing

TL;DR

DiffMorpher introduces a novel method for smooth image morphing using diffusion models by interpolating LoRA parameters and latent noises, achieving superior results over previous techniques.

Contribution

It is the first approach to enable natural image interpolation with diffusion models through semantic-aware LoRA and noise interpolation.

Findings

Achieves significantly better morphing effects than previous methods.

Effectively captures semantic transitions without annotations.

Demonstrates versatility across various object categories.

Abstract

Diffusion models have achieved remarkable image generation quality surpassing previous generative models. However, a notable limitation of diffusion models, in comparison to GANs, is their difficulty in smoothly interpolating between two image samples, due to their highly unstructured latent space. Such a smooth interpolation is intriguing as it naturally serves as a solution for the image morphing task with many applications. In this work, we present DiffMorpher, the first approach enabling smooth and natural image interpolation using diffusion models. Our key idea is to capture the semantics of the two images by fitting two LoRAs to them respectively, and interpolate between both the LoRA parameters and the latent noises to ensure a smooth semantic transition, where correspondence automatically emerges without the need for annotation. In addition, we propose an attention interpolation and injection technique and a new sampling schedule to further enhance the smoothness between consecutive images. Extensive experiments demonstrate that DiffMorpher achieves starkly better image morphing effects than previous methods across a variety of object categories, bridging a critical functional gap that distinguished diffusion models from GANs.