NSTL国家科技图书文献中心

1. MixtureGrowth: Growing Neural Networks by Recombining Learned Parameters NSTL国家科技图书文献中心

Chau Pham | Piotr Teterwak ... - 《2024 IEEE/CVF Winter Conference on Applications of Computer Vision: WACV 2024, Waikoloa, Hawaii, USA, 3-8 January 2024, [v.11]》 - IEEE/CVF Winter Conference on Applications of Computer Vision - 2024, - 2788～2797 - 共10页

摘要：Most deep neural networks are trained under fixed network architectures and require retraining when the architecture changes. If expanding the network’s size is needed, it is necessary to retrain from...

关键词： Computer vision | Codes | Computational modeling | Noise | Training data | Computer architecture | Artificial neural networks

2. Learning to Compose SuperWeights for Neural Parameter Allocation Search NSTL国家科技图书文献中心

Piotr Teterwak | Soren Nelson ... - 《2024 IEEE/CVF Winter Conference on Applications of Computer Vision: WACV 2024, Waikoloa, Hawaii, USA, 3-8 January 2024, [v.11]》 - IEEE/CVF Winter Conference on Applications of Computer Vision - 2024, - 2739～2748 - 共10页

摘要：Neural parameter allocation search (NPAS) automates parameter sharing by obtaining weights for a network given an arbitrary, fixed parameter budget. Prior work has two major drawbacks we aim to addres...

关键词： Training | Weight measurement | Computer vision | Computational modeling | Computer architecture | Network architecture | Size measurement

3. Tune it the Right Way: Unsupervised Validation of Domain Adaptation via Soft Neighborhood Density NSTL国家科技图书文献中心

Kuniaki Saito | Donghyun Kim ... - 《2021 IEEE/CVF International Conference on Computer Vision: 18th IEEE/CVF International Conference on Computer Vision (ICCV), 11-17 Oct. 2021, Virtual Event, Montreal, QC, Canada》 - International Conference on Computer Vision - 2021, - 9164～9173 - 共10页

摘要：Unsupervised domain adaptation (UDA) methods can dramatically improve generalization on unlabeled target domains. However, optimal hyper-parameter selection is critical to achieving high accuracy and ...

关键词： Training | Image segmentation | Computer vision | Codes | Density measurement | Computational modeling | Semantics

NSTL主题词： tuning | Domain | Road

4. OCONet: Image Extrapolation by Object Completion NSTL国家科技图书文献中心

Richard Strong Bowen | Huiwen Chang ... - 《2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition: CVPR 2021, Virtual Conference, 19-25 June 2021, [v.1]》 - IEEE/CVF Conference on Computer Vision and Pattern Recognition - 2021, - 2307～2317 - 共11页

摘要：Image extrapolation extends an input image beyond the originally-captured field of view. Existing methods struggle to extrapolate images with salient objects in the foreground or are limited to very s...

关键词： Extrapolation | Computer vision | Semantics | Computer architecture | Pattern recognition

NSTL主题词： extrapolation method | Completion | Image | images | Imagery (Psychotherapy)

5. Boundless: Generative Adversarial Networks for Image Extension NSTL国家科技图书文献中心

Dilip Krishnan | Piotr Teterwak ... - 《2019 IEEE/CVF International Conference on Computer Vision: IEEE/CVF International Conference on Computer Vision (ICCV 2019), 27 Oct.-2 Nov. 2019, Seoul, Korea》 - International Conference on Computer Vision - 2019, - 10520～10529 - 共10页

摘要：Image extension models have broad applications in image editing, computational photography and computer graphics. While image inpainting has been extensively studied in the literature, it is challengi...

关键词： Gallium nitride | Semantics | Generators | Training | Generative adversarial networks | Image edge detection | Context modeling

NSTL主题词： Image | images | Network

6. Understanding Invariance via Feedforward Inversion of Discriminatively Trained Classifiers NSTL国家科技图书文献中心

Piotr Teterwak | Chiyuan Zhang ... - 《International Conference on Machine Learning: ICML 2021, Online, 18-24 July 2021, Part 13 of 16》 - International Conference on Machine Learning - 2022, - 10215～10225 - 共11页

摘要：A discriminatively trained neural net classifier can fit the training data perfectly if all information about its input other than class membership has been discarded prior to the output layer. Surpri...

NSTL主题词： invariance | FEEDFORWARD | Classifiers | Inversion | Comprehension

7. OCONet: Image Extrapolation by Object Completion NSTL国家科技图书文献中心

Richard Strong Bowen | Huiwen Chang ... - 《2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition: CVPR 2021, Virtual Conference, 19-25 June 2021, [v.4]》 - IEEE/CVF Conference on Computer Vision and Pattern Recognition - 2021, - 2307～2317 - 共11页

摘要：Image extrapolation extends an input image beyond the originally-captured field of view. Existing methods struggle to extrapolate images with salient objects in the foreground or are limited to very s...

8. Boundless: Generative Adversarial Networks for Image Extension NSTL国家科技图书文献中心

Piotr Teterwak | Aaron Sarna ... - 《2019 IEEE/CVF International Conference on Computer Vision: ICCV 2019, Seoul, Korea, 27 October - 2 November 2019, [v.17]》 - IEEE/CVF International Conference on Computer Vision - 2019, - 10520～10529 - 共10页

摘要：Image extension models have broad applications in image editing, computational photography and computer graphics. While image inpainting has been extensively studied in the literature, it is challengi...

摘要：Most deep neural networks are trained under fixed network architectures and require retraining when the architecture changes. If expanding the network’s size is needed, it is necessary to retrain from...

关键词： Computer vision | Codes | Computational modeling | Noise | Training data | Computer architecture | Artificial neural networks

摘要：Neural parameter allocation search (NPAS) automates parameter sharing by obtaining weights for a network given an arbitrary, fixed parameter budget. Prior work has two major drawbacks we aim to addres...

关键词： Training | Weight measurement | Computer vision | Computational modeling | Computer architecture | Network architecture | Size measurement

摘要：Unsupervised domain adaptation (UDA) methods can dramatically improve generalization on unlabeled target domains. However, optimal hyper-parameter selection is critical to achieving high accuracy and ...

关键词： Training | Image segmentation | Computer vision | Codes | Density measurement | Computational modeling | Semantics

NSTL主题词： tuning | Domain | Road

摘要：Image extrapolation extends an input image beyond the originally-captured field of view. Existing methods struggle to extrapolate images with salient objects in the foreground or are limited to very s...

关键词： Extrapolation | Computer vision | Semantics | Computer architecture | Pattern recognition

NSTL主题词： extrapolation method | Completion | Image | images | Imagery (Psychotherapy)

摘要：Image extension models have broad applications in image editing, computational photography and computer graphics. While image inpainting has been extensively studied in the literature, it is challengi...

关键词： Gallium nitride | Semantics | Generators | Training | Generative adversarial networks | Image edge detection | Context modeling

NSTL主题词： Image | images | Network

摘要：A discriminatively trained neural net classifier can fit the training data perfectly if all information about its input other than class membership has been discarded prior to the output layer. Surpri...

NSTL主题词： invariance | FEEDFORWARD | Classifiers | Inversion | Comprehension

摘要：Image extrapolation extends an input image beyond the originally-captured field of view. Existing methods struggle to extrapolate images with salient objects in the foreground or are limited to very s...

摘要：Image extension models have broad applications in image editing, computational photography and computer graphics. While image inpainting has been extensively studied in the literature, it is challengi...

NSTL主题词： Image | images | Imagery (Psychotherapy) | Network

4008-161-200 800-990-8900

国家科技图书文献中心