中文
相关论文

相关论文: Existence and Estimation of Critical Batch Size fo…

200 篇论文

Adversarial training (AT) has been demonstrated to be effective in improving model robustness by leveraging adversarial examples for training. However, most AT methods are in face of expensive time and computational cost for calculating…

计算机视觉与模式识别 · 计算机科学 2022-07-13 Xiaojun Jia , Yong Zhang , Baoyuan Wu , Jue Wang , Xiaochun Cao

In the future, powerful AI systems may be deployed in high-stakes settings, where a single failure could be catastrophic. One technique for improving AI safety in high-stakes settings is adversarial training, which uses an adversary to…

Generative adversarial networks or GANs are a type of generative modeling framework. GANs involve a pair of neural networks engaged in a competition in iteratively creating fake data, indistinguishable from the real data. One notable…

计算机视觉与模式识别 · 计算机科学 2021-06-17 Eric J. Nunn , Pejman Khadivi , Shadrokh Samavi

Despite the dramatic success in image generation, Generative Adversarial Networks (GANs) still face great challenges in synthesizing sequences of discrete elements, in particular human language. The difficulty in generator training arises…

计算与语言 · 计算机科学 2023-02-24 Yekun Chai , Qiyue Yin , Junge Zhang

Even though deep neural networks succeed on many different tasks including semantic segmentation, they lack on robustness against adversarial examples. To counteract this exploit, often adversarial training is used. However, it is known…

计算机视觉与模式识别 · 计算机科学 2021-07-01 Daniel Wiens , Barbara Hammer

Generative Adversarial Networks (GANs) have proven to be a powerful framework for learning to draw samples from complex distributions. However, GANs are also notoriously difficult to train, with mode collapse and oscillations a common…

机器学习 · 统计学 2018-11-28 Kevin J Liang , Chunyuan Li , Guoyin Wang , Lawrence Carin

Over the past few years, there has been growing interest in developing larger and deeper neural networks, including deep generative models like generative adversarial networks (GANs). However, GANs typically come with high computational…

机器学习 · 计算机科学 2023-11-21 Yite Wang , Jing Wu , Naira Hovakimyan , Ruoyu Sun

Generative Adversarial Networks have been employed successfully to generate high-resolution augmented images of size 1024^2. Although the augmented images generated are unprecedented, the training time of the model is exceptionally high.…

图像与视频处理 · 电气工程与系统科学 2022-02-28 Atharva Karwande , Pranesh Kulkarni , Tejas Kolhe , Akshay Joshi , Soham Kamble

This paper introduces a promising alternative method for training Generative Adversarial Networks (GANs) on large-scale datasets with clear theoretical guarantees. GANs are typically learned through a minimax game between a generator and a…

计算机视觉与模式识别 · 计算机科学 2025-01-22 Chang Wan , Ke Fan , Xinwei Sun , Yanwei Fu , Minglu Li , Yunliang Jiang , Zhonglong Zheng

In adversarial learning, discriminator often fails to guide the generator successfully since it distinguishes between real and generated images using silly or non-robust features. To alleviate this problem, this brief presents a simple but…

机器学习 · 计算机科学 2021-01-20 Yong-Goo Shin , Yoon-Jae Yeo , Sung-Jea Ko

Training large-scale models under given resources requires careful design of parallelism strategies. In particular, the efficiency notion of critical batch size (CBS), concerning the compromise between time and compute, marks the threshold…

机器学习 · 计算机科学 2025-04-22 Hanlin Zhang , Depen Morwani , Nikhil Vyas , Jingfeng Wu , Difan Zou , Udaya Ghai , Dean Foster , Sham Kakade

Despite being impactful on a variety of problems and applications, the generative adversarial nets (GANs) are remarkably difficult to train. This issue is formally analyzed by \cite{arjovsky2017towards}, who also propose an alternative…

计算机视觉与模式识别 · 计算机科学 2018-03-06 Xiang Wei , Boqing Gong , Zixia Liu , Wei Lu , Liqiang Wang

Deploying models on target domain data subject to distribution shift requires adaptation. Test-time training (TTT) emerges as a solution to this adaptation under a realistic scenario where access to full source domain data is not available,…

机器学习 · 计算机科学 2023-03-21 Yongyi Su , Xun Xu , Tianrui Li , Kui Jia

Active learning parallelization is widely used, but typically relies on fixing the batch size throughout experimentation. This fixed approach is inefficient because of a dynamic trade-off between cost and speed -- larger batches are more…

Fine-tuning large language models on new data improves task performance but degrades capabilities learned during pretraining, a phenomenon known as catastrophic forgetting. Existing methods mitigate this by modifying the fine-tuning…

机器学习 · 计算机科学 2026-05-20 Parjanya Prajakta Prashant , Jiongli Zhu , Aldan Creo , Babak Salimi

The typical training of neural networks using large stepsize gradient descent (GD) under the logistic loss often involves two distinct phases, where the empirical risk oscillates in the first phase but decreases monotonically in the second…

机器学习 · 统计学 2024-06-28 Yuhang Cai , Jingfeng Wu , Song Mei , Michael Lindsey , Peter L. Bartlett

Scalability has driven recent advances in generative modeling, yet its principles remain underexplored for adversarial learning. We investigate the scalability of Generative Adversarial Networks (GANs) through two design choices that have…

计算机视觉与模式识别 · 计算机科学 2026-05-27 Sangeek Hyun , MinKyu Lee , Jae-Pil Heo

In this paper we propose a conditioning trick, called difference departure from normality, applied on the generator network in response to instability issues during GAN training. We force the generator to get closer to the departure from…

A fundamental challenge in Deep Learning is to find optimal step sizes for stochastic gradient descent automatically. In traditional optimization, line searches are a commonly used method to determine step sizes. One problem in Deep…

机器学习 · 计算机科学 2022-02-22 Maximus Mutschler , Kevin Laube , Andreas Zell

Generative adversarial networks (GANs) are successful deep generative models. GANs are based on a two-player minimax game. However, the objective function derived in the original motivation is changed to obtain stronger gradients when…

机器学习 · 统计学 2016-11-10 Masatoshi Uehara , Issei Sato , Masahiro Suzuki , Kotaro Nakayama , Yutaka Matsuo