Zobrazeno 1 - 10
of 6 668
pro vyhledávání: '"Tran Minh, A."'
Noisy labels pose a substantial challenge in machine learning, often resulting in overfitting and poor generalization. Sharpness-Aware Minimization (SAM), as demonstrated in Foret et al. (2021), improves generalization over traditional Stochastic Gra
Externí odkaz:
http://arxiv.org/abs/2411.17132
Data-Free Knowledge Distillation (DFKD) is an advanced technique that enables knowledge transfer from a teacher model to a student model without relying on original training data. While DFKD methods have achieved success on smaller datasets like CIFA
Externí odkaz:
http://arxiv.org/abs/2411.17046
In this paper we investigate the existence, uniqueness and exponential stability of pseudo almost periodic (PAP-) mild solutions of the parabolic-elliptic (P-E) Keller-Segel system on a bounded domain $\Omega\in \mathbb{R}^n$ with smooth boundary. Fi
Externí odkaz:
http://arxiv.org/abs/2411.06341
Amodal Instance Segmentation (AIS) presents an intriguing challenge, including the segmentation prediction of both visible and occluded parts of objects within images. Previous methods have often relied on shape prior information gleaned from trainin
Externí odkaz:
http://arxiv.org/abs/2409.18256
Autor:
Bui, Nhat-Tan, Hoang, Dinh-Hieu, Trinh, Quoc-Huy, Tran, Minh-Triet, Nguyen, Truong, Gauch, Susan
Negation is a fundamental linguistic concept used by humans to convey information that they do not desire. Despite this, there has been minimal research specifically focused on negation within vision-language tasks. This lack of research means that v
Externí odkaz:
http://arxiv.org/abs/2409.06481
Quantum error mitigation (QEM) can recover accurate expectation values from a noisy quantum computer by trading off bias for variance, such that an averaged result is more accurate but takes longer to converge. Probabilistic error cancellation (PEC)
Externí odkaz:
http://arxiv.org/abs/2409.04401
A long memory and non-linear realized volatility model class is proposed for direct Value at Risk (VaR) forecasting. This model, referred to as RNN-HAR, extends the heterogeneous autoregressive (HAR) model, a framework known for efficiently capturing
Externí odkaz:
http://arxiv.org/abs/2408.13588
In this work we revisit the existence, uniqueness and exponential decay of some classes of mild solutions which are almost periodic (AP-), asymptotically almost periodic (AAP-) and pseudo almost periodic (PAP-) of the scalar Lin\'eard equation by emp
Externí odkaz:
http://arxiv.org/abs/2408.12094
Autor:
Staffilani, Gigliola, Tran, Minh-Binh
Inspired by the pioneering work of Escobedo and Velazquez \cite{EscobedoVelazquez:2015:FTB,EscobedoVelazquez:2015:OTT}, we prove that solutions of 4-wave kinetic equations, under very general forms of dispersion relations, develop condensation at the
Externí odkaz:
http://arxiv.org/abs/2407.18533
On the energy transfer towards large values of wavenumbers for solutions of 4-wave kinetic equations
Autor:
Staffilani, Gigliola, Tran, Minh-Binh
Inspired by the fundamental work of Escobedo and Velazquez \cite{EscobedoVelazquez:2015:FTB,EscobedoVelazquez:2015:OTT}, we prove that solutions of 4-wave kinetic equations, under very general forms of the dispersion relations, exhibit the transfer o
Externí odkaz:
http://arxiv.org/abs/2407.18508