Filters








38,803 Hits in 5.8 sec

Budget-Aware Adapters for Multi-Domain Learning [article]

Rodrigo Berriel, Stéphane Lathuilière, Moin Nabi, Tassilo Klein, Thiago Oliveira-Santos, Nicu Sebe, Elisa Ricci
2019 arXiv   pre-print
To this aim, we introduce Budget-Aware Adapters that select the most relevant feature channels to better handle data from a novel domain.  ...  Multi-Domain Learning (MDL) refers to the problem of learning a set of models derived from a common deep architecture, each one specialized to perform a task in a certain domain (e.g., photos, sketches  ...  Acknowledgements This work was carried out under the "Vision and Learning joint Laboratory" between FBK and UNITN, and financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior  ... 
arXiv:1905.06242v2 fatcat:ao2g3vo6ibb4xfaw463flnxnam

Budget-Aware Adapters for Multi-Domain Learning

Rodrigo Berriel, Stephane Lathuillere, Moin Nabi, Tassilo Klein, Thiago Oliveira-Santos, Nicu Sebe, Elisa Ricci
2019 2019 IEEE/CVF International Conference on Computer Vision (ICCV)  
To this aim, we introduce Budget-Aware Adapters that select the most relevant feature channels to better handle data from a novel domain.  ...  Multi -Domain Learning (MDL) refers to the problem of learning a set of models derived from a common deep architecture, each one specialized to perform a task in a certain domain (e.g., photos, sketches  ...  Acknowledgements This work was carried out under the "Vision and Learning joint Laboratory" between FBK and UNITN, and financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior  ... 
doi:10.1109/iccv.2019.00047 dblp:conf/iccv/BerrielLNKOS019 fatcat:wvt272llafhvda6sskwrp33cia

Communication-Cost Aware Microphone Selection For Neural Speech Enhancement with Ad-hoc Microphone Arrays [article]

Jonah Casebeer, Jamshed Kaikaus, Paris Smaragdis
2021 arXiv   pre-print
In this paper, we present a method for jointly-learning a microphone selection mechanism and a speech enhancement network for multi-channel speech enhancement with an ad-hoc microphone array.  ...  When communication cost is not an issue, both beamforming and deep learning techniques have proven successful for multi-channel speech enhancement.  ...  For the i th microphone, the encoder outputs z[t] (i) given the time-domain input chunk x i [t · h : t · h + f ].  ... 
arXiv:2011.07348v4 fatcat:uegfkjgoqjaoniwtwe6rch4ary

Self-Path: Self-supervision for Classification of Pathology Images with Limited Annotations [article]

Navid Alemi Koohbanani, Balagopal Unnikrishnan, Syed Ali Khurram, Pavitra Krishnaswamy, Nasir Rajpoot
2020 arXiv   pre-print
We introduce novel domain specific self-supervision tasks that leverage contextual, multi-resolution and semantic features in pathology images for semi-supervised learning and domain adaptation.  ...  Further, we show that Self-Path improves domain adaptation for classification of histology image patches when there is no labeled data available for the target domain.  ...  : Domain prediction can enable representation learning for domain adaptation.  ... 
arXiv:2008.05571v1 fatcat:tbgp42venreotlofe2fien4jhu

Shared resource aware scheduling on power-constrained tiled many-core processors

Sudhanshu Shekhar Jha, Wim Heirman, Ayose Falcón, Jordi Tubella, Antonio González, Lieven Eeckhout
2016 Proceedings of the ACM International Conference on Computing Frontiers - CF '16  
This calls for hierarchical solutions, such as on-chip voltage regulators per-tile rather than per-core, along with multi-level power management.  ...  . • Shared DVFS and cache adaptation can degrade performance of co-scheduled threads on a tile. • DVFS and cache-aware thread migration (DCTM) to ensure optimum per-tile co-scheduling of compatible threads  ...  In our tiled architecture, we assume each core's micro-architecture can be adapted individually. DVFS. DVFS adaptation is a widely used technique for enforcing power budgeting.  ... 
doi:10.1145/2903150.2903490 dblp:conf/cd/JhaHFT0E16 fatcat:v7uohhnrm5cw5ne5o5c3ut2yvu

Shared resource aware scheduling on power-constrained tiled many-core processors

Sudhanshu Shekhar Jha, Wim Heirman, Ayose Falcón, Jordi Tubella, Antonio González, Lieven Eeckhout
2017 Journal of Parallel and Distributed Computing  
This calls for hierarchical solutions, such as on-chip voltage regulators per-tile rather than per-core, along with multi-level power management.  ...  . • Shared DVFS and cache adaptation can degrade performance of co-scheduled threads on a tile. • DVFS and cache-aware thread migration (DCTM) to ensure optimum per-tile co-scheduling of compatible threads  ...  In our tiled architecture, we assume each core's micro-architecture can be adapted individually. DVFS. DVFS adaptation is a widely used technique for enforcing power budgeting.  ... 
doi:10.1016/j.jpdc.2016.10.001 fatcat:tepv7eywendzjifk7foewvgbsi

NASIB: Neural Architecture Search withIn Budget [article]

Abhishek Singh, Anubhav Garg, Jinan Zhou, Shiv Ram Dubey, Debo Dutta
2019 arXiv   pre-print
In this paper, we propose a new approach for NAS, called NASIB, which adapts and attunes to the computation resources (budget) available by varying the exploration vs. exploitation trade-off.  ...  The proposed method can provide the architecture search useful for different computation resources and different domains beyond image classification of natural images where we lack bespoke architecture  ...  c) applying architecture search to different domains or performing multi-objective architecture search.  ... 
arXiv:1910.08665v1 fatcat:5s3solxtjjhddo32up665fnmrm

Online Budgeted Learning for Classifier Induction [article]

Eran Fainman, Bracha Shapira, Lior Rokach, Yisroel Mirsky
2019 arXiv   pre-print
In this paper we introduce the issue of online budgeted learning and describe a general framework for addressing this challenge.  ...  Our comparative study on five real-world datasets indicates that adaptive policies outperform random policies for most budget limitations and datasets.  ...  When data acquisition is limited by budget, intelligent feature acquisition becomes a necessity, and this domain is referred to as budgeted learning.  ... 
arXiv:1903.05382v1 fatcat:iea5rkzod5acnng3qaj6p4jwhu

A Survey on Practical Applications of Multi-Armed and Contextual Bandits [article]

Djallel Bouneffouf, Irina Rish
2019 arXiv   pre-print
Specifically, we introduce a taxonomy of common MAB-based applications and summarize state-of-art for each of those domains.  ...  performance combined with certain attractive properties, such as learning from less feedback.  ...  In [Huo and Fu, 2017] , the authors incorporate risk-awareness into the classic multi-armed bandit setting and introduce a novel algorithm for portfolio construction.  ... 
arXiv:1904.10040v1 fatcat:j6v37wy7f5bmvpfzzhtnutbeoa

Leadership Institute for Faculty Development

Douglas J. Gould, Michelle Hammond
2021 Medical Science Educator  
Resources provided N/A April 1, 2, 3 Multi-domain leadership workshop 2-h workshop on multi-domain leadership 3 April-May 1 360-degree assessment and coaching Completion of custom 360 evaluation  ...  Likewise, self-awareness is a key driver of leadership development and awareness is especially important in adaption, transition, and changes [11] .  ... 
doi:10.1007/s40670-020-01179-6 pmid:33495719 pmcid:PMC7817139 fatcat:gtfvnvbcgbcsvd2zs3wdxju4de

Anycost GANs for Interactive Image Synthesis and Editing [article]

Ji Lin, Richard Zhang, Frieder Ganz, Song Han, Jun-Yan Zhu
2021 arXiv   pre-print
Anycost GAN can be executed at various cost budgets (up to 10x computation reduction) and adapt to a wide range of hardware and latency requirements.  ...  By using sampling-based multi-resolution training, adaptive-channel training, and a generator-conditioned discriminator, the anycost generator can be evaluated at various configurations while achieving  ...  We thank Taesung Park and Zhixin Shu for the helpful discussion. Part of the work is supported under NSF CAREER Award #1943349. We thank MIT-IBM Watson AI Lab for the support.  ... 
arXiv:2103.03243v1 fatcat:jw5nbgxqmze67ljg6s5o4kuo4i

End-to-End Entity Resolution for Big Data: A Survey [article]

Vassilis Christophides, Vasilis Efthymiou, Themis Palpanas, George Papadakis, Kostas Stefanidis
2020 arXiv   pre-print
One of the most important tasks for improving data quality and the reliability of data analytics results is Entity Resolution (ER).  ...  We present the basic concepts, processing steps and execution strategies that have been proposed by different communities, i.e., database, semantic Web and machine learning, in order to cope with the loose  ...  Meta-blocking has been adapted to both multi-core [134] and MapReduce parallelization [55] .  ... 
arXiv:1905.06397v3 fatcat:rs2qoolz2jcppklriew5pjfefq

Should We Be Pre-training? An Argument for End-task Aware Training as an Alternative [article]

Lucio M. Dery, Paul Michel, Ameet Talwalkar, Graham Neubig
2022 arXiv   pre-print
We next introduce an online meta-learning algorithm that learns a set of multi-task weights to better balance among our multiple auxiliary objectives, achieving further improvements on end-task performance  ...  On three different low-resource NLP tasks from two domains, we demonstrate that multi-tasking the end-task and auxiliary objectives results in significantly better downstream task performance than the  ...  Domain and Task Adaptive Pre-training Gururangan et al. (2020) present Domain Adaptive Pre-Training (DAPT) and Task Adaptive Pre-Training (TAPT) as methods for continued pre-training.  ... 
arXiv:2109.07437v2 fatcat:ehuvvggh2ba7lfs6jeg6w3ss3u

Multidimensional Belief Quantification for Label-Efficient Meta-Learning [article]

Deep Pandey, Qi Yu
2022 arXiv   pre-print
We propose a novel uncertainty-aware task selection model for label efficient meta-learning.  ...  However, learning from few samples introduces uncertainty, and quantifying model confidence for few-shot predictions is essential for many critical domains.  ...  We would also like to thank the anonymous reviewers for their constructive comments.  ... 
arXiv:2203.12768v1 fatcat:spzkdqejhfbwnijstgyachbybe

How to Reach Real-Time AI on Consumer Devices? Solutions for Programmable and Custom Architectures [article]

Stylianos I. Venieris and Ioannis Panopoulos and Ilias Leontiadis and Iakovos S. Venieris
2021 arXiv   pre-print
Moreover, we showcase how custom ASIC- and FPGA-based accelerators can be an enabling factor for next-generation AI applications, such as multi-DNN systems.  ...  Collectively, these results highlight the critical need for further exploration as to how the various cross-stack solutions can be best combined in order to bring the latest advances in deep learning close  ...  Primary assumption in both cases is the availability of the training dataset for the target AI task, which enables the model-level modifications. 1) Hardware-aware Model Adaptation: Recently, a plethora  ... 
arXiv:2106.15021v1 fatcat:b25jifosajeuba57qxiaockmg4
« Previous Showing results 1 — 15 out of 38,803 results