Search Results - "Iwana, Brian Kenji"

Time series classification using local distance-based features in multi-modal fusion networks by Kenji Iwana, Brian, Uchida, Seiichi

“…•Proposes a method of using the local distances between time series as features.•Uses multi-modal CNNs to combine the raw coordinates and local distance…”
Get full text

An empirical survey of data augmentation for time series classification with neural networks by Iwana, Brian Kenji, Uchida, Seiichi

“…In recent times, deep artificial neural networks have achieved many successes in pattern recognition. Part of this success can be attributed to the reliance on…”
Get full text

Guided neural style transfer for shape stylization by Atarsaikhan, Gantugs, Iwana, Brian Kenji, Uchida, Seiichi

“…Designing logos, typefaces, and other decorated shapes can require professional skills. In this paper, we aim to produce new and unique decorated shapes by…”
Get full text

Learning the micro deformations by max-pooling for offline signature verification by Zheng, Yuchen, Iwana, Brian Kenji, Malik, Muhammad Imran, Ahmed, Sheraz, Ohyama, Wataru, Uchida, Seiichi

“…•Proving that CNNs have the potential to capture “micro deformations”.•Learning the discrimination from a two-phase CNN based feature extractor.•Achieving…”
Get full text

Mining the displacement of max-pooling for text recognition by Zheng, Yuchen, Iwana, Brian Kenji, Uchida, Seiichi

“…•A new feature named “displacement features” are extracted from the pooling layer.•Combining the displacement features and pooling features for text…”
Get full text

On Mini-Batch Training with Varying Length Time Series by Kenji Iwana, Brian

“…In real-world time series recognition applications, it is possible to have data with varying length patterns. However, when using artificial neural networks…”
Get full text

FETNet: Feature erasing and transferring network for scene text removal by Lyu, Guangtao, Liu, Kun, Zhu, Anna, Uchida, Seiichi, Iwana, Brian Kenji

“…•We propose a novel FETNet which could remove scene text near completely in images.•Our method is formulated in a one stage way and is trained in an end to end…”
Get full text

Few-Shot Text Style Transfer via Deep Feature Similarity by Zhu, Anna, Lu, Xiongbo, Bai, Xiang, Uchida, Seiichi, Iwana, Brian Kenji, Xiong, Shengwu

“…Generating text to have a consistent style with only a few observed highly-stylized text samples is a difficult task for image processing. The text style…”
Get full text

Tunable U-Net: Controlling Image-to-Image Outputs Using a Tunable Scalar Value by Kang, Seokjun, Uchida, Seiichi, Iwana, Brian Kenji

“…Image-to-image conversion tasks are more accurate and sophisticated than ever thanks to advances in deep learning. However, since typical deep learning models…”
Get full text

Dynamic Weight Alignment for Temporal Convolutional Neural Networks by Iwana, Brian Kenji, Uchida, Seiichi

“…In this paper, we propose a method of improving temporal Convolutional Neural Networks (CNN) by determining the optimal alignment of weights and inputs using…”
Get full text

How do Convolutional Neural Networks Learn Design? by Jolly, Shailza, Iwana, Brian Kenji, Kuroki, Ryohei, Uchida, Seiichi

“…In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique…”
Get full text

Explaining Convolutional Neural Networks using Softmax Gradient Layer-wise Relevance Propagation by Iwana, Brian Kenji, Kuroki, Ryohei, Uchida, Seiichi

“…Convolutional Neural Networks (CNN) have become state-of-the-art in the field of image classification. However, not everything is understood about their inner…”
Get full text

On Mini-Batch Training with Varying Length Time Series by Iwana, Brian Kenji

“…In real-world time series recognition applications, it is possible to have data with varying length patterns. However, when using artificial neural networks…”
Get full text

Complex image processing with less data—Document image binarization by integrating multiple pre-trained U-Net modules by Kang, Seokjun, Iwana, Brian Kenji, Uchida, Seiichi

“…•Propose a novel document binarization method by cascading pre-trained U-Nets.•Use pre-trained U-Net for solving a training image shortage problem.•Study for…”
Get full text

FETNet: Feature Erasing and Transferring Network for Scene Text Removal by Lyu, Guangtao, Liu, Kun, Zhu, Anna, Uchida, Seiichi, Iwana, Brian Kenji

“…Pattern Recognition 2023 The scene text removal (STR) task aims to remove text regions and recover the background smoothly in images for private information…”
Get full text

DTW-NN: A novel neural network for time series recognition using dynamic alignment between inputs and weights by Iwana, Brian Kenji, Frinken, Volkmar, Uchida, Seiichi

“…This paper describes a novel model for time series recognition called a Dynamic Time Warping Neural Network (DTW-NN). DTW-NN is a feedforward neural network…”
Get full text

Time Series Data Augmentation for Neural Networks by Time Warping with a Discriminative Teacher by Iwana, Brian Kenji, Uchida, Seiichi

“…Neural networks have become a powerful tool in pattern recognition and part of their success is due to generalization from using large datasets. However,…”
Get full text

On the Ability of a CNN to Realize Image-to-Image Language Conversion by Baba, Kohei, Uchida, Seiichi, Iwana, Brian Kenji

“…The purpose of this paper is to reveal the ability that Convolutional Neural Networks (CNN) have on the novel task of image-to-image language conversion. We…”
Get full text

Deep Dynamic Time Warping: End-to-End Local Representation Learning for Online Signature Verification by Wu, Xiaomeng, Kimura, Akisato, Iwana, Brian Kenji, Uchida, Seiichi, Kashino, Kunio

“…Siamese networks have been shown to be successful in learning deep representations for multivariate time series verification. However, most related studies…”
Get full text

What is the Reward for Handwriting? - A Handwriting Generation Model Based on Imitation Learning by Kanda, Keisuke, Iwana, Brian Kenji, Uchida, Seiichi

“…Analyzing the handwriting generation process is an important issue and has been tackled by various generation models, such as kinematics based models and…”
Get full text