Pseudoinverse-Guided Diffusion Models for Inverse Problems

Authors: Jiaming Song, Arash Vahdat, Morteza Mardani, Jan Kautz

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our method, termed Pseudoinverse-Guided Diffusion Models (ΠGDM), on various inverse problems, such as super-resolution, inpainting, and JPEG restoration over Image Net validation images, and show that it achieves similar performance when compared against state-of-the-art taskspecific diffusion models (Saharia et al., 2021; Dhariwal & Nichol, 2021; Saharia et al., 2022a).
Researcher Affiliation -1 Anonymous authors Paper under double-blind review
Pseudocode Yes We list the full algorithm for ΠGDM for VP-SDE in Algorithm 1. ... Listing 1: Pseudocode for computing the pseudoinverse guidance for the noiseless case.
Open Source Code No The paper refers to publicly available datasets and model checkpoints from 'openai/guided-diffusion' used in their experiments, but does not explicitly state that their own implementation code for ΠGDM is open-source or provide a link to it.
Open Datasets Yes We evaluate quantitative results on the Image Net dataset (Russakovsky et al., 2015)
Dataset Splits Yes We report super-resolution results on the full Image Net validation set, and to follow the earlier practice established in Saharia et al. (2022a), we report inpainting and JPEG restoration results on a subset that contains 10k images3.
Hardware Specification No The paper does not specify the hardware (e.g., GPU models, CPU types) used for running the experiments.
Software Dependencies No The paper mentions 'Py Torch-like implementation' but does not specify any software dependencies with version numbers (e.g., specific library versions or programming language versions).
Experiment Setup Yes We use 100 iterations and η = 1.0 for ΠGDM, and include additional task-specific details in App. B. For ΠGDM, we use a class-conditional model, initialize our sampler from pure Gaussian noise at the maximum noise level σT , apply 100 iterations to each image, and set η = 1.0.