Summary of I2am: Interpreting Image-to-image Latent Diffusion Models Via Bi-attribution Maps, by Junseo Park and Hyeryung Jang
I2AM: Interpreting Image-to-Image Latent Diffusion Models via Bi-Attribution Maps
by Junseo Park, Hyeryung Jang
First submitted to arxiv on: 17 Jul 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper introduces Image-to-Image Attribution Maps (I2AM), a novel method to enhance the interpretability of image-to-image (I2I) diffusion models. I2AM visualizes bidirectional attribution maps, from the reference image to the generated image and vice versa, by aggregating cross-attention scores across time steps, attention heads, and layers. This allows for insights into how critical features are transferred between images. The method is demonstrated on object detection, inpainting, and super-resolution tasks, showing that I2AM successfully identifies key regions responsible for generating the output even in complex scenes. Additionally, a novel evaluation metric, Inpainting Mask Attention Consistency Score (IMACS), is introduced to assess the alignment between attribution maps and inpainting masks, which correlates strongly with existing performance metrics. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper explains how it introduces a new method called Image-to-Image Attribution Maps (I2AM) that helps us understand why image-to-image diffusion models make certain choices. It shows how I2AM works by looking at how features are transferred between images and identifies important regions for generating the output. The method is tested on different tasks like object detection, filling in missing parts of an image, and making a low-resolution image look higher quality. |
Keywords
» Artificial intelligence » Alignment » Attention » Cross attention » Diffusion » Mask » Object detection » Super resolution
