Zobrazeno 1 - 10
of 12
pro vyhledávání: '"Aysegul Dundar"'
Autor:
Sukru Burc Eryilmaz, Aysegul Dundar
Publikováno v:
IEEE Transactions on Neural Networks and Learning Systems
We analyze why the orthogonality penalty improves quantization in deep neural networks. Using results from perturbation theory as well as through extensive experiments with Resnet50, Resnet101, and VGG19 models, we mathematically and experimentally s
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ac16b634863257f860d445397b0207fb
https://doi.org/10.1109/tnnls.2022.3171297
https://doi.org/10.1109/tnnls.2022.3171297
Autor:
Guilin Liu, Aysegul Dundar, Kevin J. Shih, Ting-Chun Wang, Fitsum A. Reda, Karan Sapra, Zhiding Yu, Xiaodong Yang, Andrew Tao, Bryan Catanzaro
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence
Partial convolution weights convolutions with binary masks and renormalizes on valid pixels. It was originally proposed for image inpainting task because a corrupted image processed by a standard convolutional often leads to artifacts. Therefore, bin
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::419b13711150547070912d3574787649
https://pubmed.ncbi.nlm.nih.gov/36155473
https://pubmed.ncbi.nlm.nih.gov/36155473
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence. 43:2360-2372
Generating computer graphics (CG) rendered synthetic images has been widely used to create simulation environments for robotics/autonomous driving and generate labeled data. Yet, the problem of training models purely with synthetic data remains chall
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a7608e09dc3d8b1e802a2649bc9d94e0
https://doi.org/10.1109/tpami.2020.2969421
https://doi.org/10.1109/tpami.2020.2969421
Publikováno v:
Lecture Notes in Computer Science ISBN: 9783031197864
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::b4a569204671e71c84a789ba1135f1da
https://doi.org/10.1007/978-3-031-19787-1_9
https://doi.org/10.1007/978-3-031-19787-1_9
Publikováno v:
CVPR
Conditional image synthesis for generating photorealistic images serves various applications for content editing to content generation. Previous conditional image synthesis algorithms mostly rely on semantic maps, and often fail in complex environmen
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6345d3c948e06e496d88be15e4a4eb55
http://arxiv.org/abs/2004.10289
http://arxiv.org/abs/2004.10289
Publikováno v:
IEEE Transactions on Neural Networks and Learning Systems. 28:1572-1583
Deep convolutional neural networks (DCNNs) have become a very powerful tool in visual perception. DCNNs have applications in autonomous robots, security systems, mobile phones, and automobiles, where high throughput of the feedforward evaluation phas
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::be272ad4cd435a4ec1d454a62f4db026
https://doi.org/10.1109/tnnls.2016.2545298
https://doi.org/10.1109/tnnls.2016.2545298
Autor:
Fitsum A. Reda, Aysegul Dundar, Guilin Liu, Bryan Catanzaro, Andrew Tao, Deqing Sun, Mohammad Shoeybi, Jan Kautz, Kevin J. Shih
Publikováno v:
ICCV
Learning to synthesize high frame rate videos via interpolation requires large quantities of high frame rate training videos, which, however, are scarce, especially at high resolutions. Here, we propose unsupervised techniques to synthesize high fram
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::7bb040d0a081e1b0eb4b813dd95104ea
https://doi.org/10.1109/iccv.2019.00098
https://doi.org/10.1109/iccv.2019.00098
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence
( Early Access ) Unsupervised landmark learning is the task of learning semantic keypoint-like representations without the use of expensive input keypoint-level annotations. A popular approach is to factorize an image into a pose and appearance data
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f490ad2edc5cf07003196637bbf712ee
https://doi.org/10.1109/tpami.2021.3055560
https://doi.org/10.1109/tpami.2021.3055560
Publikováno v:
HPEC
In this paper, we present a memory access optimized routing scheme for a hardware accelerated real-time implementation of deep convolutional neural networks (DCNNs) on a mobile platform. DCNNs consist of multiple layers of 3D convolutions, each compr
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::db659aa1b483e313f9a1330d37b8832e
https://doi.org/10.1109/hpec.2014.7040963
https://doi.org/10.1109/hpec.2014.7040963