Výsledky vyhledávání - "Vongkulbhisal, Jayakorn"

Report

A Two-Block RNN-based Trajectory Prediction from Incomplete Trajectory

Autor: Fujii, Ryo, Vongkulbhisal, Jayakorn, Hachiuma, Ryo, Saito, Hideo

Trajectory prediction has gained great attention and significant progress has been made in recent years. However, most works rely on a key assumption that each video is successfully preprocessed by detection and tracking algorithms and the complete o

Externí odkaz: http://arxiv.org/abs/2203.07098

Zobrazit plný text záznamu

Report

Ensemble of Discriminators for Domain Adaptation in Multiple Sound Source 2D Localization

Autor: Moing, Guillaume Le, Agravante, Don Joven, Inoue, Tadanobu, Vongkulbhisal, Jayakorn, Munawar, Asim, Tachibana, Ryuki, Vinayavekhin, Phongtharin

This paper introduces an ensemble of discriminators that improves the accuracy of a domain adaptation technique for the localization of multiple sound sources. Recently, deep neural networks have led to promising results for this task, yet they requi

Externí odkaz: http://arxiv.org/abs/2012.05908

Zobrazit plný text záznamu

Report

Data-Efficient Framework for Real-world Multiple Sound Source 2D Localization

Autor: Moing, Guillaume Le, Vinayavekhin, Phongtharin, Agravante, Don Joven, Inoue, Tadanobu, Vongkulbhisal, Jayakorn, Munawar, Asim, Tachibana, Ryuki

Deep neural networks have recently led to promising results for the task of multiple sound source localization. Yet, they require a lot of training data to cover a variety of acoustic conditions and microphone array layouts. One can leverage acoustic

Externí odkaz: http://arxiv.org/abs/2012.05533

Zobrazit plný text záznamu

Report

Learning Multiple Sound Source 2D Localization

Autor: Moing, Guillaume Le, Vinayavekhin, Phongtharin, Inoue, Tadanobu, Vongkulbhisal, Jayakorn, Munawar, Asim, Tachibana, Ryuki, Agravante, Don Joven

In this paper, we propose novel deep learning based algorithms for multiple sound source localization. Specifically, we aim to find the 2D Cartesian coordinates of multiple sound sources in an enclosed environment by using multiple microphone arrays.

Externí odkaz: http://arxiv.org/abs/2012.05515

Zobrazit plný text záznamu

Report

On Focal Loss for Class-Posterior Probability Estimation: A Theoretical Perspective

Autor: Charoenphakdee, Nontawat, Vongkulbhisal, Jayakorn, Chairatanakul, Nuttapong, Sugiyama, Masashi

The focal loss has demonstrated its effectiveness in many real-world applications such as object detection and image classification, but its theoretical understanding has been limited so far. In this paper, we first prove that the focal loss is class

Externí odkaz: http://arxiv.org/abs/2011.09172

Zobrazit plný text záznamu

Report

Unifying Heterogeneous Classifiers with Distillation

Autor: Vongkulbhisal, Jayakorn, Vinayavekhin, Phongtharin, Visentini-Scarzanella, Marco

In this paper, we study the problem of unifying knowledge from a set of classifiers with different architectures and target classes into a single classifier, given only a generic set of unlabelled data. We call this problem Unifying Heterogeneous Cla

Externí odkaz: http://arxiv.org/abs/1904.06062

Zobrazit plný text záznamu

Report

Discriminative Optimization: Theory and Applications to Computer Vision Problems

Autor: Vongkulbhisal, Jayakorn, De la Torre, Fernando, Costeira, João P.

Publikováno v: IEEE Transactions on Pattern Analysis and Machine Intelligence ( Volume: 41, Issue: 4, Apr 2019 )

Many computer vision problems are formulated as the optimization of a cost function. This approach faces two main challenges: (i) designing a cost function with a local optimum at an acceptable solution, and (ii) developing an efficient numerical met

Externí odkaz: http://arxiv.org/abs/1707.04318

Zobrazit plný text záznamu