% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@INPROCEEDINGS{Cherti:905638,
author = {Cherti, Mehdi and Jitsev, Jenia},
title = {{E}ffect of {P}re-{T}raining {S}cale on {I}ntra- and
{I}nter-{D}omain {F}ull and {F}ew-{S}hot {T}ransfer
{L}earning for {N}atural and {M}edical {X}-{R}ay {C}hest
{I}mages},
reportid = {FZJ-2022-00865},
pages = {1-6},
year = {2021},
abstract = {Transfer learning aims to exploit pre-trained models for
more efficient follow-up training on wide range of
downstream tasks and datasets, enabling successful training
also on small data. Recently, strong improvement was shown
for transfer learning and model generalization when
increasing model, data and compute budget scale in the
pre-training. To compare effect of scale both in intra- and
inter-domain full and few-shot transfer, in this study we
combine for the first time large openly available medical
X-Ray chest imaging datasets to reach a dataset scale
comparable to ImageNet-1k. We then conduct pre-training and
transfer to different natural or medical targets while
varying network size and source data scale and domain, being
either large natural (ImageNet-1k/21k) or large medical
chest X-Ray datasets. We observe strong improvement due to
larger pre-training scale for intra-domain natural-natural
and medical-medical transfer. For inter-domain
natural-medical transfer, we find improvements due to larger
pre-training scale on larger X-Ray targets in full shot
regime, while for smaller targets and for few-shot regime
the improvement is not visible. Remarkably, large networks
pre-trained on very large natural ImageNet-21k are as good
or better than networks pre-trained on largest available
medical X-Ray data when performing transfer to large X-Ray
targets. We conclude that high quality models for
inter-domain transfer can be also obtained by substantially
increasing scale of model and generic natural source data,
removing necessity for large domain-specific medical source
data in the pre-training. Code is available at:
$\url{https://github.com/SLAMPAI/large-scale-pretraining-transfer}}$},
month = {Dec},
date = {2021-12-06},
organization = {Medical Imaging Meets NeurIPS
(MedNeurIPS), Sydney / online
(Australia), 6 Dec 2021 - 14 Dec 2021},
cin = {JSC},
cid = {I:(DE-Juel1)JSC-20090406},
pnm = {5112 - Cross-Domain Algorithms, Tools, Methods Labs (ATMLs)
and Research Groups (POF4-511)},
pid = {G:(DE-HGF)POF4-5112},
typ = {PUB:(DE-HGF)8},
eprint = {2106.00116},
howpublished = {arXiv:2106.00116},
archivePrefix = {arXiv},
SLACcitation = {$\%\%CITATION$ = $arXiv:2106.00116;\%\%$},
url = {https://juser.fz-juelich.de/record/905638},
}