Takio Kurita

I'm a researcher in computer vision and machine learning at Hiroshima University currently working on deep learning applications and pattern recognition. I previously did research at ETL (Electrotechnical Laboratory) and have been involved with various collaborative projects across multiple institutions. My work focuses on developing novel deep learning architectures and improving existing machine learning techniques for visual recognition tasks. My research encompasses several key areas including anomaly detection using autoencoders, facial expression recognition, and neural network optimization. I have extensive experience with support vector machines, kernel methods, and dimensionality reduction techniques like principal component analysis. Recently, I've been exploring advanced topics such as neural radiance fields (NeRF), vision transformers, and self-supervised learning approaches. I'm particularly interested in developing robust methods that can handle noisy data and limited training samples.
Throughout my career, I've published extensively on topics ranging from traditional pattern recognition methods to cutting-edge deep learning techniques. My work spans both theoretical contributions and practical applications, including medical image analysis, autonomous driving systems, and multimedia database retrieval. I continue to investigate how to make machine learning models more reliable and interpretable, with recent focus on attention mechanisms and explainable AI approaches.

Publications

A Practical Application of Machine Learning for the Development of Metallole-Based Fluorescent Materials

Yusuke Kanematsu, Akiyoshi Ohta, Shunya Nagai, Yohei Adachi, Hiromasa Kaneko, Takayoshi Ishimoto, Takio Kurita, J. Ohshita

Molecules 2025

ABS PDF

Direction-aware convolutional autoencoder based on positional encoding for one-dimensional anomaly detection

Qien Yu, Qiong Chang, Tinghui Ouyang, Takio Kurita, Ran Dong

Information Sciences 2025

Takio Kurita

Publications

A Practical Application of Machine Learning for the Development of Metallole-Based Fluorescent Materials

Direction-aware convolutional autoencoder based on positional encoding for one-dimensional anomaly detection

Nonlinear dimensionality reduction with q-Gaussian distribution

Improved Head and Data Augmentation to Reduce Artifacts at Grid Boundaries in Object Detection

An Object Detection Method Using Probability Maps for Instance Segmentation to Mask Background

Facial Image Manipulation via Discriminative Decomposition of Semantic Space

Neural Radiance Fields with Regularizer Based on Differences of Neighboring Pixels

Convolutional autoencoder based on latent subspace projection for anomaly detection.

Data Expansion Approach with Attention Mechanism for Learning with Noisy Labels

Single-Image Super-Resolution Reconstruction Based on the Differences of Neighboring Pixels

Sample Selection Approach with Number of False Predictions for Learning with Noisy Labels

Graph Laplacian Regularization based on the Differences of Neighboring Pixels for Conditional Convolutions for Instance Segmentation

Deep Neural Network Models for Colon Cancer Screening

Summary of Works on Image Classification with Noisy Labels

Pixel Relationships-based Regularizer for Retinal Vessel Image Segmentation

Attention-effective multiple instance learning on weakly stem cell colony segmentation

Consistency Regularization on Clean Samples for Learning with Noisy Labels

Invariant Feature Extraction for CNN Classifier by using Gradient Reversal Layer

Leveraging network using controlled weight learning approach for thyroid cancer lymph node detection

Regularizer based on Euler characteristic for retinal blood vessel segmentation

Correction to: Extensive framework based on novel convolutional and variational autoencoder based on maximization of mutual information for anomaly detection

Parametric q-Gaussian distributed stochastic neighbor embedding with Convolutional Neural Network

Deep Metric Learning for Multi-Label and Multi-Object Image Retrieval

Extensive framework based on novel convolutional and variational autoencoder based on maximization of mutual information for anomaly detection

Autoencoder framework based on orthogonal projection constraints improves anomalies detection

Convolutional Neural Network based on Temporal Pose Features for Surgical Procedure Recognition

Noise Reduction of SEM Images using U-net with SSIM Loss Function

Weakly-Supervised Action Localization, and Action Recognition Using Global–Local Attention of 3D CNN

q-SNE: Visualizing Data using q-Gaussian Distributed Stochastic Neighbor Embedding

Mixture of experts with convolutional and variational autoencoders for anomaly detection

Channel Planting for Deep Neural Networks using Knowledge Distillation

Filter Pruning using Hierarchical Group Sparse Regularization for Deep Convolutional Neural Networks

Non-negative Matrix Factorization of a set of Economic Time Series with Graph Based Smoothing of Basis Vectors and Sparseness of the Coefficients

Completion of Missing Labels for Multi-Label Annotation by a Unified Graph Laplacian Regularization

Author Correction: Deep learning enables automated localization of the metastatic lymph node for thyroid cancer on 131I post-ablation whole-body planar scans

Robust pruning for efficient CNNs

Deep learning enables automated localization of the metastatic lymph node for thyroid cancer on 131I post-ablation whole-body planar scans

Triplet Loss for Knowledge Distillation

Adaptive Neuron-wise Discriminant Criterion and Adaptive Center Loss at Hidden Layer for Deep Convolutional Neural Network

Hierarchical Group Sparse Regularization for Deep Convolutional Neural Networks

Analysis of Information Flow in Hidden Layers of the Trained Neural Network by Canonical Correlation Analysis

Short-Term Action Recognition by 3D Convolutional Neural Network with Pixel-Wise Evidences

Transfer Learning by Cascaded Network to Identify and Classify Lung Nodules for Cancer Detection

On-line non-overlapping camera calibration net

Prediction of Sequential Organelles Localization under Imbalance using A Balanced Deep U-Net

Siamese Network for Classification with Optimization of AUC

Learning with Incomplete Labels for Multi-label Image Annotation Using CNN and Restricted Boltzmann Machines

U-Net with Graph Based Smoothing Regularizer for Small Vessel Segmentation on Fundus Image

Detection of One Dimensional Anomalies Using a Vector-Based Convolutional Autoencoder

Multi instance learning via deep CNN for multi-class recognition of Alzheimer's disease

Hilbert Vector Convolutional Neural Network: 2D Neural Network on 1D Data

Automatic Segmentation of Mandibular Cortical Bone on Cone-Beam CT Images Based on Histogram Thresholding and Polynomial Fitting

Video Super Resolution with Estimation of Motion Information by Using Higher Resolution Images Obtained by Single Image Super Resolution

Multi-projection deep learning network for segmentation of 3D medical images

Segmenting Tooth Components in Dental X-Ray Images Using Gaussian Kernel- Based Conditional Spatial Fuzzy C-Means Clustering Algorithm

Sparse discriminant analysis based on estimation of posterior probabilities

Mixup of Feature Maps in a Hidden Layer for Training of Convolutional Neural Network

Convolutional Neural Network with Discriminant Criterion for Input of Each Neuron in Output Layer

3D Region Merging for Segmentation of Teeth on Cone-Beam Computed Tomography Images

Facial expression intensity estimation using Siamese and triplet networks

Motion Superpixels for Temporal Video Classification

Texture Segmentation using Siamese Network and Hierarchical Region Merging

Correlation Net : spatio temporal multimodal deep learning

Deep Packet Flow: Action Recognition via Multiresolution Deep Wavelet Packet of Local Dense Optical Flows

Logistic Regression Analysis for the Material Design of Chiral Crystals

Critical texture pattern feature assessment for characterizing colonies of induced pluripotent stem cells through machine learning techniques

Marker-based non-overlapping camera calibration methods with additional support camera views

Data Visualization for Deep Neural Networks Based on Interlayer Canonical Correlation Analysis

Deep vector-based convolutional neural network approach for automatic recognition of colonies of induced pluripotent stem cells

Gated spatio and temporal convolutional neural network for activity recognition: towards gated multimodal deep learning

Image Inpainting by Recursive Estimation Using Neural Network and Wavelet Transformation

Fast and Accurate Image Super Resolution by Deep CNN with Skip Connection and Network in Network

Tree-Wise Discriminative Subtree Selection for Texture Image Labeling

Deep CNN with Graph Laplacian Regularization for Multi-label Image Annotation

Effect of Additive Noise for Multi-Layered Perceptron with AutoEncoders

Temporal Evolution of Motion Superpixel for Video Classification

Consecutive Dimensionality Reduction by Canonical Correlation Analysis for Visualization of Convolutional Neural Networks

Improvement of learning for CNN with ReLU activation by sparse regularization