Explicitly Minimizing the Blur Error of Variational Autoencoders

Authors: Gustav Bredell, Kyriakos Flouris, Krishna Chaitanya, Ertunc Erdil, Ender Konukoglu

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

Reproducibility Variable Result LLM Response
Research Type Experimental We show the potential of the proposed loss on three different data sets, where it outperforms several recently proposed reconstruction losses for VAEs.
Researcher Affiliation Academia Gustav Bredell, Kyriakos Flouris, Krishna Chaitanya, Ertunc Erdil & Ender Konukoglu Department of Information Technology and Electrical Engineering ETH Zurich gustav.bredell@vision.ee.ethz.ch
Pseudocode No The paper describes the approach using mathematical formulations and textual descriptions but does not include any clearly labeled pseudocode or algorithm blocks.
Open Source Code No Details on the architecture can be found in the code that will be publicly available upon acceptance.
Open Datasets Yes To evaluate the potential of the proposed approach on natural images we make use of the popular Celeb A dataset as provided by Liu et al. (2015) and Lee et al. (2020) for low and high-resolution images, respectively.
Dataset Splits No For all the datasets we use 80% for the training set and 20% for the test set.
Hardware Specification No The paper does not specify any particular hardware components (e.g., CPU, GPU models, or memory size) used for running the experiments.
Software Dependencies No The code is written in Python and Py Torch (Paszke et al. (2019)) is used as library for the deep learning models.
Experiment Setup Yes Furthermore, the Adam optimizer is used with a learning rate of 1e 4. For the lowand high-resolution Celeb A the number of training epochs were, 100 and 200, respectively. For the HCP dataset the models were trained for 400 epochs.