From 253d39f29bdd46fe38596fea10ce0d2b66a78f51 Mon Sep 17 00:00:00 2001 From: Lyndon Duong Date: Tue, 10 Jan 2023 10:16:55 -0500 Subject: [PATCH] edit authors and text --- joss_paper/paper.md | 41 ++++++++++++++++++++++------------------- 1 file changed, 22 insertions(+), 19 deletions(-) diff --git a/joss_paper/paper.md b/joss_paper/paper.md index c2d5c72d..9e03d0a3 100644 --- a/joss_paper/paper.md +++ b/joss_paper/paper.md @@ -1,5 +1,5 @@ --- -title: 'Plenoptic: synthesis methods for analyzing model representations' +title: 'Plenoptic.py: Synthesizing model-optimized visual stimuli' tags: - Python - PyTorch @@ -10,7 +10,7 @@ authors: - name: Kathryn Bonnen orcid: 0000-0002-9210-8275 affiliation: 1, 2 - - name: William Broderick + - name: William F. Broderick orcid: 0000-0002-8999-9003 affiliation: 1 - name: Lyndon R. Duong @@ -22,6 +22,12 @@ authors: - name: Nikhil Parthasarathy orcid: 0000-0003-2572-6492 affiliation: 1 + - name: Xinyuan Zhao + orcid: 0000-0003-2572-6492 + affiliation: 1 + - name: Thomas E. Yerxa + orcid: 0000-0003-2572-6492 + affiliation: 1 - name: Eero P. Simoncelli orcid: 000-0002-1206-527X affiliation: 1, 2 @@ -30,32 +36,30 @@ affiliations: index: 1 - name: Center for Computational Neuroscience, Flatiron Institute, New York, NY, USA index: 2 -date: April 2021 +date: January 2023 bibliography: references.bib --- # Summary - -``Plenoptic`` builds primarily off of ``PyTorch`` [@paszke_pytorch_2019], a Python machine learning library popular in the research community due to its rapid prototyping capability. With ``Plenoptic``, users can build and train models in ``PyTorch``, then use ``Plenoptic`` synthesis methods to assess their internal representations. -Our library is easily extensible, and allows for great flexibility to those who wish to develop or test their own synthesis methods. -Within the library, we also provide an extensive suite of ``PyTorch``-implemented models and activation functions canonical to computational neuroscience. - -Many of the methods in ``Plenoptic`` have been developed and used across several studies; however, analyses in these studies used disparate languages and frameworks, and some have yet to be made publicly available. -Here, we have reimplemented the methods central to each of these studies, and unified them under a single, fully-documented API. -Our library includes several Jupyter notebook tutorials designed to be accessible to researchers in the fields of machine learning, and computational neuroscience, and perceptual science. -``Plenoptic`` provides an exciting avenue for researchers to probe their models to gain a deeper understanding of their internal representations. - -# Statement of Need - -# Overview +In sensory perception and neuroscience, new computational models are most often tested and compared in terms of their ability to fit existing data sets. +However, experimental data are inherently limited in size, quality, and type, and complex models often saturate their explainable variance. +Moreover, it is often difficult to use models to guide the development of future experiments. +Here, building on ideas for optimal experimental stimulus selection (e.g., QUEST, Watson and Pelli, 1983), we present "Plenoptic", a python software library for generating visual stimuli optimized for testing or comparing models. +Plenoptic provides a unified framework containing four previously-published synthesis methods -- model metamers (Freeman and Simoncelli, 2011), Maximum Differentiation (MAD) competition (Wang and Simoncelli, 2008), eigen-distortions (Berardino et al. 2017), and representational geodesics (Hénaff and Simoncelli, 2015) -- each of which offers visualization of model representations, and generation of images that can be used to experimentally test alignment with the human visual system. +Plenoptic leverages modern machine-learning methods to enable application of these synthesis methods to any computational model that satisfies a small set of common requirements. +The most important of these is that the model must be image-computable, implemented in PyTorch, and end-to-end differentiable. +The package includes examples of several low- and mid-level visual models, as well as a set of perceptual quality metrics. +Plenoptic is open source, tested, documented, and extensible, allowing the broader research community to contribute new examples and methods. +In summary, Plenoptic leverages machine learning tools to tighten the scientific hypothesis-testing loop, facilitating investigation of human visual representations. # Acknowledgements -KB, WB, LRD, PEF, and NP each contributed equally to this work; and names are listed alphabetically. -EPS was funded by the Howard Hughes Medical Institute. EPS and KB were funded by Simons Institute. +All authors contributed equally to this work; and names are listed alphabetically. +EPS and KB were funded by Simons Institute. For a quick reference, the following citation commands can be used: + - `@author:2001` -> "Author et al. (2001)" - `[@author:2001]` -> "(Author et al., 2001)" - `[@author1:2001; @author2:2001]` -> "(Author1 et al., 2001; Author2 et al., 2002)" @@ -69,4 +73,3 @@ For a quick reference, the following citation commands can be used: @wang_maximum_2008 @paszke_pytorch_2019 @portilla_parametric_2000 -