Enhancing Conditional Image Generation with Explainable Latent Space Manipulation

TL;DR

This paper introduces a novel diffusion-based image synthesis method that uses explainable latent space manipulation and gradient-based attention to improve fidelity to reference images and conditional prompts, achieving superior results in quality and alignment.

Contribution

The paper presents a new approach combining diffusion models with explainable latent space manipulation and gradient-based attention, enhancing fidelity and controllability in image generation.

Findings

Achieves lower FID scores indicating better image fidelity.

Demonstrates high CLIP scores for text-image alignment.

Outperforms baseline models in quality metrics.

Abstract

In the realm of image synthesis, achieving fidelity to a reference image while adhering to conditional prompts remains a significant challenge. This paper proposes a novel approach that integrates a diffusion model with latent space manipulation and gradient-based selective attention mechanisms to address this issue. Leveraging Grad-SAM (Gradient-based Selective Attention Manipulation), we analyze the cross attention maps of the cross attention layers and gradients for the denoised latent vector, deriving importance scores of elements of denoised latent vector related to the subject of interest. Using this information, we create masks at specific timesteps during denoising to preserve subjects while seamlessly integrating the reference image features. This approach ensures the faithful formation of subjects based on conditional prompts, while concurrently refining the background for a more coherent composition. Our experiments on places365 dataset demonstrate promising results, with our proposed model achieving the lowest mean and median Frechet Inception Distance (FID) scores compared to baseline models, indicating superior fidelity preservation. Furthermore, our model exhibits competitive performance in aligning the generated images with provided textual descriptions, as evidenced by high CLIP scores. These results highlight the effectiveness of our approach in both fidelity preservation and textual context preservation, offering a significant advancement in text-to-image synthesis tasks.