Filters








945 Hits in 15.0 sec

On the Transfer of Inductive Bias from Simulation to the Real World: a New Disentanglement Dataset [article]

Muhammad Waleed Gondal and Manuel Wüthrich and ore Miladinović and Francesco Locatello and Martin Breidt and Valentin Volchkov and Joel Akpo and Olivier Bachem and Bernhard Schölkopf and Stefan Bauer
<span title="2019-11-25">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
These datasets provide for the first time the possibility to systematically investigate how well different disentanglement methods perform on real data in comparison to simulation, and how simulated data  ...  Since real-world data is notoriously costly to collect, many recent state-of-the-art disentanglement models have heavily relied on synthetic toy data-sets.  ...  We would also like to thank Felix Grimminger, Ludovic Righetti, Stefan Schaal, Julian Viereck and Felix Widmaier whose work served as a starting point for the development of the robotic platform in the  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1906.03292v3">arXiv:1906.03292v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/esytaew4cje5bmpxlavn6ifuy4">fatcat:esytaew4cje5bmpxlavn6ifuy4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201009211907/https://arxiv.org/pdf/1906.03292v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/35/e0/35e0531d547a3b25db948969a2181de9a79586b4.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1906.03292v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Shallow Bayesian Meta Learning for Real-World Few-Shot Recognition [article]

Xueting Zhang, Debin Meng, Henry Gouk, Timothy Hospedales
<span title="2021-08-27">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
It is agnostic to the off-the-shelf features chosen and thus will continue to benefit from advances in feature representations.  ...  Empirically, it leads to robust performance in cross-domain few-shot learning and, crucially for real-world applications, it leads to better uncertainty calibration in predictions.  ...  Finally, by meta-learning the prior, we are able to optimize inductive bias for few-shot learning performance. Both LDA and QDA benefit from meta-learning, but QDA performs better overall.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.02833v2">arXiv:2101.02833v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/zckpwqurtvdvnnoqy55ugayygy">fatcat:zckpwqurtvdvnnoqy55ugayygy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210901222407/https://arxiv.org/pdf/2101.02833v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ba/6a/ba6a3dbc04caa1ee17b82a79b66452a4ac619b55.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.02833v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

ROBIN : A Benchmark for Robustness to Individual Nuisances in Real-World Out-of-Distribution Shifts [article]

Bingchen Zhao, Shaozuo Yu, Wufei Ma, Mingxin Yu, Shenxiao Mei, Angtian Wang, Ju He, Alan Yuille, Adam Kortylewski
<span title="2021-12-02">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this work, we introduce ROBIN, a benchmark dataset for diagnosing the robustness of vision algorithms to individual nuisances in real-world images.  ...  ROBIN builds on 10 rigid categories from the PASCAL VOC 2012 and ImageNet datasets and includes out-of-distribution examples of the objects 3D pose, shape, texture, context and weather conditions.  ...  To close the gap between the performance of vision models on datasets and the performance in the real-world, many techniques has been proposed.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.14341v2">arXiv:2111.14341v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/2o24o5mdzvdyrnuwlcbnatd3wa">fatcat:2o24o5mdzvdyrnuwlcbnatd3wa</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211204101941/https://arxiv.org/pdf/2111.14341v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/6f/bd/6fbdaeac87f6ae18c82eccb998c5a0800a798109.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.14341v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Mastering Atari with Discrete World Models [article]

Danijar Hafner, Timothy Lillicrap, Mohammad Norouzi, Jimmy Ba
<span title="2022-02-12">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
DreamerV2 constitutes the first agent that achieves human-level performance on the Atari benchmark of 55 tasks by learning behaviors inside a separately trained world model.  ...  We introduce DreamerV2, a reinforcement learning agent that learns behaviors purely from predictions in the compact latent space of a powerful world model.  ...  Acknowledgements We thank our anonymous reviewers for their feedback and Nick Rhinehart for an insightful discussion about the potential benefits of categorical latent variables.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.02193v4">arXiv:2010.02193v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/7rjazcce75e7tbky5gpemuaqbq">fatcat:7rjazcce75e7tbky5gpemuaqbq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210507000944/https://arxiv.org/pdf/2010.02193v3.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/c2/27/c227b55f17f2753ff6e6fa8db3829d78d2b89f42.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.02193v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

The Role of Pretrained Representations for the OOD Generalization of Reinforcement Learning Agents [article]

Andrea Dittadi, Frederik Träuble, Manuel Wüthrich, Felix Widmaier, Peter Gehler, Ole Winther, Francesco Locatello, Olivier Bachem, Bernhard Schölkopf, Stefan Bauer
<span title="2022-04-16">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Building sample-efficient agents that generalize out-of-distribution (OOD) in real-world settings remains a fundamental unsolved problem on the path towards achieving higher-level cognition.  ...  By training 240 representations and over 10,000 reinforcement learning (RL) policies on a simulated robotic setup, we evaluate to what extent different properties of pretrained VAE-based representations  ...  Part of the experiments were generously supported with compute credits by Amazon Web Services.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.05686v4">arXiv:2107.05686v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/my5n552v4ncdxmrqikhhapngzy">fatcat:my5n552v4ncdxmrqikhhapngzy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220307190022/https://arxiv.org/pdf/2107.05686v3.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/47/2f/472f7ab75d260a2b4c4e1b04bced7adda2aebd40.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.05686v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Beyond Flatland: Pre-training with a Strong 3D Inductive Bias [article]

Shubhaankar Gupta, Thomas P. O'Connell, Bernhard Egger
<span title="2021-11-30">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Since the image formation process in the natural world is based on its 3D structure, we expect pre-training with 3D mesh renders to provide an implicit bias leading to better generalization capabilities  ...  Pre-training on large-scale databases consisting of natural images and then fine-tuning them to fit the application at hand, or transfer-learning, is a popular strategy in computer vision.  ...  Better transfer performance after pre-training on any of our 3D object datasets relative to the 2D fractal datasets would support our hypothesis that pre-training with a strong 3D inductive bias will more  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.00113v1">arXiv:2112.00113v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/n7lecrbnkfhj3hlceaobkkjmti">fatcat:n7lecrbnkfhj3hlceaobkkjmti</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211203044505/https://arxiv.org/pdf/2112.00113v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/32/b4/32b4cd119a625c4f0c2cdb3b83aef7a0a6de3c58.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.00113v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

On Disentangled Representations Learned From Correlated Data [article]

Frederik Träuble, Elliot Creager, Niki Kilbertus, Francesco Locatello, Andrea Dittadi, Anirudh Goyal, Bernhard Schölkopf, Stefan Bauer
<span title="2021-07-16">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this work, we bridge the gap to real-world scenarios by analyzing the behavior of the most prominent disentanglement approaches on correlated data in a large-scale empirical study (including 4260 models  ...  The focus of disentanglement approaches has been on identifying independent factors of variation in data.  ...  On the transfer of inductive bias from simulation to the real world: a new disentanglement dataset. In Advances in Neural Information Processing Systems, 2019.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.07886v3">arXiv:2006.07886v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/irkinvphx5drdep6fe2kul2iby">fatcat:irkinvphx5drdep6fe2kul2iby</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210721005210/https://arxiv.org/pdf/2006.07886v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ab/f8/abf8c40e1c8bc0f02db9c69163ab2844679205ab.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.07886v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Learning Disentangled Representations in the Imaging Domain [article]

Xiao Liu, Pedro Sanchez, Spyridon Thermos, Alison Q. O'Neil, Sotirios A. Tsaftaris
<span title="2022-04-17">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Disentangled representation learning has been proposed as an approach to learning general representations even in the absence of, or with limited, supervision.  ...  A good general representation can be fine-tuned for new target tasks using modest amounts of data, or used directly in unseen domains achieving remarkable performance in the corresponding task.  ...  We thank the participants of the DREAM tutorials for feedback.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2108.12043v5">arXiv:2108.12043v5</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/cbpmp6pbajhjvjzovulswuj2wy">fatcat:cbpmp6pbajhjvjzovulswuj2wy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211114111221/https://arxiv.org/pdf/2108.12043v4.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/41/b7/41b73eb430f9aa44bf20913cafa43f2f429b4e69.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2108.12043v5" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence [article]

Nicholas Roy, Ingmar Posner, Tim Barfoot, Philippe Beaudoin, Yoshua Bengio, Jeannette Bohg, Oliver Brock, Isabelle Depatie, Dieter Fox, Dan Koditschek, Tomas Lozano-Perez, Vikash Mansinghka (+8 others)
<span title="2021-10-28">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In reality, therefore, these limitations result in learning-based systems which suffer from many of the same operational shortcomings as more traditional, engineering-based approaches when deployed on  ...  Consequently, the notion of applying learning methods to a particular problem set has become an established and valuable modus operandi to advance a particular field.  ...  Acknowledgments This paper is the results of a three day workshop sponsored by and held at Element AI, and their support is very gratefully acknowledged.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.15245v1">arXiv:2110.15245v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/juxc4tai2jbklpul55loccnp7e">fatcat:juxc4tai2jbklpul55loccnp7e</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211031113309/https://arxiv.org/pdf/2110.15245v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/9c/d8/9cd80a501ea586086f614f08667e7ed7533a1204.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.15245v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Disentangling Factors of Variation Using Few Labels [article]

Francesco Locatello, Michael Tschannen, Stefan Bauer, Gunnar Rätsch, Bernhard Schölkopf, Olivier Bachem
<span title="2020-02-14">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We observe that a small number of labeled examples (0.01--0.5\% of the data set), with potentially imprecise and incomplete labels, is sufficient to perform model selection on state-of-the-art unsupervised  ...  In this paper, we investigate the impact of such supervision on state-of-the-art disentanglement methods and perform a large scale study, training over 52000 models under well-defined and reproducible  ...  Acknowledgments: Francesco Locatello is supported by the Max Planck ETH Center for Learning Systems, by an ETH core grant (to Gunnar Rätsch), and by a Google Ph.D. Fellowship.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1905.01258v2">arXiv:1905.01258v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/3nvwprvr4vhc5jrl2lycls7sg4">fatcat:3nvwprvr4vhc5jrl2lycls7sg4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200321125734/https://arxiv.org/pdf/1905.01258v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1905.01258v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Disentangled Spatiotemporal Graph Generative Models [article]

Yuanqi Du and Xiaojie Guo and Hengning Cao and Yanfang Ye and Liang Zhao
<span title="2022-02-28">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Qualitative and quantitative experiments on both synthetic and real-world datasets demonstrate the superiority of the proposed model over the state-of-the-arts by up to 69.2% for graph generation and 41.5%  ...  A variational objective function and new mutual information thresholding algorithms driven by information bottleneck theory have been proposed to maximize the disentanglement among the factors with theoretical  ...  Acknowledgement This work was supported by the National Science Foundation (NSF) Grant No. 1755850, No. 1841520, No. 2007716, No. 2007976, No. 1942594, No. 1907805, a Jeffress Memorial Trust Award, Amazon  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2203.00411v1">arXiv:2203.00411v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/wzkmje7u65dxzlstk2wt6glklu">fatcat:wzkmje7u65dxzlstk2wt6glklu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220516143648/https://arxiv.org/pdf/2203.00411v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/6c/34/6c349eeb05c55e9172e970f804166d3f986b08e3.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2203.00411v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

An Image is Worth More Than a Thousand Words: Towards Disentanglement in the Wild [article]

Aviv Gabbay, Niv Cohen, Yedid Hoshen
<span title="2021-10-25">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Unsupervised disentanglement has been shown to be theoretically impossible without inductive biases on the models and the data.  ...  While annotating the true generative factors is only required for a limited number of observations, we argue that it is infeasible to enumerate all the factors of variation that describe a real-world image  ...  While the order of the values can be exploited as an inductive bias to disentanglement, many attributes in the real world (e.g. human gender or animal specie) are not ordered in a meaningful way.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2106.15610v2">arXiv:2106.15610v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/242rvtvzjrh6zf7tmvvbmaglxm">fatcat:242rvtvzjrh6zf7tmvvbmaglxm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210701092701/https://arxiv.org/pdf/2106.15610v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/33/0d/330dac37a7d661ad8b2861810adc008b30864893.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2106.15610v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Disentangling 3D Prototypical Networks For Few-Shot Concept Learning [article]

Mihir Prabhudesai, Shamit Lal, Darshan Patil, Hsiao-Yu Tung, Adam W Harley, Katerina Fragkiadaki
<span title="2021-07-20">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
current state-of-the-art, and enable a visual question answering system that uses them as its modules to generalize one-shot to novel objects in the scene.  ...  Our networks incorporate architectural biases that reflect the image formation process, 3D geometry of the world scene, and shape-style interplay.  ...  adaptive instance normalization (Huang & Belongie, 2017) as an inductive bias to do style transfer between a pair of images.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.03367v3">arXiv:2011.03367v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/bqxpkyamcnf53cvbgrravlgk4m">fatcat:bqxpkyamcnf53cvbgrravlgk4m</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210723031357/https://arxiv.org/pdf/2011.03367v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/9e/9b/9e9be84ee3793f9bc951b7680368d1b8ec9113a2.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.03367v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Abstracts from the XIth World DOHaD Congress

<span title="">2019</span> <i title="Cambridge University Press (CUP)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/4c4pgwqxbrccxdhbjipoi2ywdu" style="color: black;">Journal of Developmental Origins of Health and Disease</a> </i> &nbsp;
The effects of paternal BMI on the odds of both SGA and LGA in male infants were similar to those of paternal height; however, paternal height Background/Aims: Children's behavioral problems are associated  ...  After explaining the objective and the process of this project to all participants, we obtained consent from them to participate in this investigation voluntarily.  ...  The Project provides real time feedback to participants on a number of health issues and encourages interventions.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1017/s2040174419000643">doi:10.1017/s2040174419000643</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/sgbbtoqodrhipjzyc33ujpt5gq">fatcat:sgbbtoqodrhipjzyc33ujpt5gq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201107043951/https://www.cambridge.org/core/services/aop-cambridge-core/content/view/4C4129DB313152B601F4DB7046CF945E/S2040174419000643a.pdf/div-class-title-abstracts-from-the-xi-span-class-sup-th-span-world-dohad-congress-div.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1017/s2040174419000643"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> cambridge.org </button> </a>

Reducing Domain Gap by Reducing Style Bias [article]

Hyeonseob Nam, HyunJae Lee, Jongchan Park, Wonjun Yoon, Donggeun Yoo
<span title="2021-04-01">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Our Style-Agnostic Networks (SagNets) disentangle style encodings from class categories to prevent style biased predictions and focus more on the contents.  ...  Recent studies suggest that one of the main causes of this problem is CNNs' strong inductive bias towards image styles (i.e. textures) which are sensitive to domain changes, rather than contents (i.e.  ...  For example, young children learn many object concepts from pictures, but they naturally transfer their knowledge to the real world [10] . * Equal contribution 1 Code: https://github.com/hyeonseobnam/  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1910.11645v4">arXiv:1910.11645v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ugujqj3jxrffjicyz5oouv5gna">fatcat:ugujqj3jxrffjicyz5oouv5gna</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210408061202/https://arxiv.org/pdf/1910.11645v4.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/f5/07/f5079d86265948177103cfd4aa23d5a0b74b39bd.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1910.11645v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 945 results