no code implementations • 22 Nov 2023 • Shoichiro Takeda, Yasunori Akagi, Naoki Marumo, Kenta Niwa
On the basis of this reduction, our algorithms solve the small optimization problem instead of the original OT.
1 code implementation • 13 Oct 2023 • Ryoma Sato, Yuki Takezawa, Han Bao, Kenta Niwa, Makoto Yamada
LLMs can generate texts that cannot be distinguished from human-written texts.
no code implementations • 2 Oct 2023 • Yuki Takezawa, Ryoma Sato, Han Bao, Kenta Niwa, Makoto Yamada
Although existing watermarking methods have successfully detected texts generated by LLMs, they significantly degrade the quality of the generated texts.
no code implementations • 30 Sep 2022 • Yuki Takezawa, Han Bao, Kenta Niwa, Ryoma Sato, Makoto Yamada
In this study, we propose Momentum Tracking, which is a method with momentum whose convergence rate is proven to be independent of data heterogeneity.
no code implementations • 23 May 2022 • Yuki Takezawa, Kenta Niwa, Makoto Yamada
However, the convergence rate of the ECL is provided only when the objective function is convex, and has not been shown in a standard machine learning setting where the objective function is non-convex.
no code implementations • 8 May 2022 • Yuki Takezawa, Kenta Niwa, Makoto Yamada
Moreover, we demonstrate that the C-ECL is more robust to heterogeneous data than the Gossip-based algorithms.
1 code implementation • 24 Mar 2022 • Guoqiang Zhang, Kenta Niwa, W. Bastiaan Kleijn
Firstly, we show that the particular placement of the parameter epsilon within the update expressions of AdaBelief reduces the range of the adaptive stepsizes, making AdaBelief closer to SGD with momentum.
no code implementations • CVPR 2022 • Shoichiro Takeda, Kenta Niwa, Mariko Isogawa, Shinya Shimizu, Kazuki Okami, Yushi Aono
Eulerian video magnification (EVM) has progressed to magnify subtle motions with a target frequency even under the presence of large motions of objects.
1 code implementation • 20 Feb 2021 • Haimin Zhang, Min Xu, Guoqiang Zhang, Kenta Niwa
We show that applying stochastic scaling at the gradient level is complementary to that applied at the feature level to improve the overall performance.
no code implementations • 21 Nov 2019 • Guo-Qiang Zhang, Kenta Niwa, W. B. Kleijn
Considering a weight matrix W from a particular neural layer in the model, our objective is to design a function h(W) such that its row vectors are approximately orthogonal to each other while allowing the DNN model to fit the training data sufficiently accurate.
no code implementations • 24 Feb 2019 • Guo-Qiang Zhang, Kenta Niwa, W. Bastiaan Kleijn
Adaptive gradient methods such as Adam have been shown to be very effective for training deep neural networks (DNNs) by tracking the second moment of gradients to compute the individual learning rates.
no code implementations • 22 Oct 2018 • Yuma Koizumi, Kenta Niwa, Yusuke Hioka, Kazunori Kobayashi, Yoichi Haneda
Since OSQA scores have been used widely for sound-quality evaluation, constructing DNNs to increase OSQA scores would be better than using the minimum-MSE to create high-quality output signals.
no code implementations • 27 Sep 2018 • Guoqiang Zhang, Kenta Niwa, W. Bastiaan Kleijn
Empirical studies for training four convolutional neural networks over MNIST and CIFAR10 show that under proper parameter selection, Game produces promising validation performance as compared to AMSGrad and PAdam.