Takio Kurita

Hiroshima University | HU · Department of Information Engineering

Linear discriminant analysis (LDA) is one of the well known methods to extract the best features for multi-class discrimination. Recently Kernel discriminant analysis (KDA) has been successfully applied in many applications. KDA is one of the nonlinear extensions of LDA and construct nonlinear discriminant mapping by using kernel functions. But the...

Linear discriminant analysis (LDA) is one of the well known methods to extract the best features for the multi-class discrimination. Otsu derived the optimal nonlinear discriminant analysis (ONDA) by assuming the underlying probabilities and showed that the ONDA was closely related to Bayesian decision theory (the posterior probabilities). Also Ots...

Visual impression may differ with each person. User-friendly interfaces for image database systems require special retrieval methods which can adapt to the visual impression of each user. Algorithms for learning personal visual impressions of visual objects are described. The algorithms are based on multivariate data analysis methods. These algorit...

This paper discusses learning algorithms of layered neural networks from the standpoint of maximum likelihood estimation. Fisher information is explicitly calculated for the network with only one neuron. It can be interpreted as a weighted covariance matrix of input vectors. A learning algorithm is presented on the basis of Fisher's scoring method....

In recent years, the dimensionality reduction has become more important as the number of dimensions of data used in various tasks such as regression and classification has increased. As popular nonlinear dimensionality reduction methods, t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP) hav...

We investigated the influence of horizontal shifts of the input images for one stage object detection method. We found that the object detector class scores drop when the target object center is at the grid boundary. Many approaches have focused on reducing the aliasing effect of down-sampling to achieve shift-invariance. However, down-sampling doe...

Image inpainting is inferring missing pixels in images using known content and surrounding regions. To achieve this task, existing methods have utilized convolutional neural networks and adversarial generative networks. However, these approaches produce blurry and unrealistic completion results when the region of the missing pixels is large and irr...

This paper proposes a two-step detector called segmented object detection, whose performance is improved by masking the background region. Previous single-stage object detection methods suffer from the problem of imbalance between foreground and background classes, where the background occupies more regions in the image than the foreground. Thus, t...

Bidirectional Generative Adversarial Networks (BiGANs) is a generative model with an invertible mapping between latent and image space. The mapping allows us to encode real images into latent representations and reconstruct input images. However, from preliminary experiments, we found that the joint probability distributions learned by the generato...

Image anomaly detection (AD) is widely researched on many occasions in computer vision tasks. High-dimensional data, such as image data, with noise and complex background is still challenging to detect anomalies under the situation that imbalanced or incomplete data are available. Some deep learning methods can be trained in an unsupervised way and...

In recent years, the development of deep learning has contributed to various areas of machine learning. However, deep learning requires a huge amount of data to train the model, and data collection techniques such as web crawling can easily generate incorrect labels. If a training dataset has noisy labels, the generalization performance of deep lea...

The task of image segmentation is to classify each pixel in the image based on the appropriate label. Various deep learning approaches have been proposed for image segmentation that offers high accuracy and deep architecture. However, the deep learning technique uses a pixel-wise loss function for the training process. Using pixel-wise loss neglect...

The deep learning technique was used to increase the performance of single image super-resolution (SISR). However, most existing CNN-based SISR approaches primarily focus on establishing deeper or larger networks to extract more significant high-level features. Usually, the pixel-level loss between the target high-resolution image and the estimated...

In recent years, deep neural networks (DNNs) have made a significant impact on a variety of research fields and applications. One drawback of DNNs is that it requires a huge amount of dataset for training. Since it is very expensive to ask experts to label the data, many non-expert data collection methods such as web crawling have been proposed. Ho...

Supervised learning has found many applications as one of the fundamental techniques in machine learning. Recently, Barlow Twins loss has been used in self-supervised learning to train Convolutional Neural Networks to extract invariant features by introducing perturbations to the input data and encouraging the network to learn features that are inv...

3D convolutional neural network (3D CNN) captures spatial and temporal information on 3D data such as video sequences. However, due to the convolution and pooling mechanism, the information loss that occurs seems unavoidable. To improve the visual explanations and classification in 3D CNN, we propose two approaches; (i) aggregate layer-wise global...

Early detection of colorectal cancer can significantly facilitate clinicians’ decision-making and reduce their workload. This can be achieved using automatic systems with endoscopic and histological images. Recently, the success of deep learning has motivated the development of image- and video-based polyp identification and segmentation. Currently...

In our brain, the visual information captured by the retina is processed by the two different visual pathways known as ventral stream and the dorsal stream. The ventral stream known as the “what pathway” is involved with object and visual identification and recognition and the dorsal stream known as “where pathway” is involved with processing the o...

In recent years, the performance of machine learning algorithms has been rapidly improved because of the progress of deep learning. To approximate any non-linear function, almost all models of deep learning use the non-linear activation function. Rectified linear units (ReLU) function is most commonly used. The continuous version of the ReLU functi...

The detection of induced pluripotent stem cell (iPSC) colonies often needs the precise extraction of the colony features. However, existing computerized systems relied on segmentation of contours by preprocessing for classifying the colony conditions were task-extensive. To maximize the efficiency in categorizing colony conditions, we propose a mul...

The neighboring relationship of the pixels is fundamental information in the image. Utilizing neighboring pixel information provides an advantage over using only pixel-to-pixel information. This study presents aregularization term based on the differences of neighboring pixels. We define the differences of neighboring pixels byusing the Graph Theor...

In the recent years, deep learning has achieved significant results in various areas of machine learning. Deep learning requires a huge amount of data to train a model, and data collection techniques such as web crawling have been developed. However, there is a risk that these data collection techniques may generate incorrect labels. If a deep lear...

This paper proposes a re-ranking method for vehicle re-identification based on k-reciprocal Encoding. In recent years, with the development and popularization of deep learning, vehicle re-identification has made great progress and excellent performance. The re-ranking me-thod applied to re-identification has been widely adopted and recognized. Howe...

The deep learning technique was used to increase the performance of single image super-resolution (SISR). However, most existing CNN-based SISR approaches primarily focus on establishing deeper or larger networks to extract more significant high-level features. Usually, the pixel-level loss between the target high-resolution image and the estimated...

In present study, we proposed a general framework based on a convolutional kernel and a variational autoencoder (CVAE) for anomaly detection on both complex image and vector datasets. The main idea is to maximize mutual information (MMI) through regularizing key information as follows: (1) the features between original input and the representation...

Identification of cervical metastatic lymph nodes (LN) on I-131 post-ablation whole-body planar scans (WBS) for cancer staging is crucial for patients with papillary thyroid cancer (PTC). The existing deep network is plagued by instability in finding the under-represented LN classes on highly complex WBS, where end-to-end tuned models that can acco...

This paper proposes an age estimation method from the age period using Triplet Network. Age estimation is still an active research topic in machine learning, and it can be formulated as a regression problem. Usually, a specific age value to each of the training face images is assigned as a correct label, and the model to estimate the age value of a...

Recent developments in technology, such as crowdsourcing and web crawling, have made it easier to train machine learning models that require big data. However, the data collected by non-experts may contain noisy labels, and training a classification model on the data will result in poor generalization performance. In particular, Deep Neural Network...

This study focused on the problem of anomaly detection (AD) by means of mixture-of-experts network. Most of the existing AD methods solely based on the reconstruction errors or latent representation using a single low-dimensional manifold are often not ideal for the image objects with complex background. However, modeling the data as a mixture of l...

Content-based image retrieval has been a hot topic among computer vision researchers for a long time. There have been many advances over the years, one of the recent ones being deep metric learning, inspired by the success of deep neural networks in many machine learning tasks. The goal of metric learning is to extract good high-level features from...

Segmentation of retinal blood vessels is important for the analysis of diabetic retinopathy (DR). Existing methods do not prioritize the small and disconnected vessels for DR. With the aim of paying attention to the small and disconnected vessel regions, this study introduced Euler characteristics (EC) from topology to calculate the number of isola...

In this study, we propose a novel autoencoder framework based on orthogonal projection constraint (OPC) for anomaly detection (AD) on both complex image and vector datasets. Orthogonal projection is useful to capture the null subspace that consists of noisy information for AD, which is explicitly ignored in the existing approaches. The exploration...

To compensate for severe shortage of scrub nurses who support surgeons during surgery, Miyawaki et al. have developed a scrub nurse robot (SNR) system [1,2,3,4,5]. One of its current challenges is how to make the SNR recognize surgical procedures which compose a surgical operation and understand/predict surgeons' intentions. Therefore, in this pape...

The image-to-image translation networks, such as U-net [1] or Pix2pix [2], are known to be able to convert input images into different images where the image quality is improved or desired semantic information hidden in the input images are extracted. Several types of research based on such image translation networks have been carried out to realiz...

3D Convolutional Neural Network (3D CNN) captures spatial and temporal information on 3D data such as video sequences. However, due to the convolution and pooling mechanism, the information loss seems unavoidable. To improve the visual explanations and classification in 3D CNN, we propose two approaches; i) aggregate layer-wise global to local (glo...

The dimensionality reduction has been widely introduced to use the high-dimensional data for regression, classification, feature analysis, and visualization. As the one technique of dimensionality reduction, a stochastic neighbor embedding (SNE) was introduced. The SNE leads powerful results to visualize high-dimensional data by considering the sim...

Since the convolutional neural networks are often trained with redundant parameters, it is possible to reduce redundant kernels or filters to obtain a compact network without dropping the classification accuracy. In this paper, we propose a filter pruning method using the hierarchical group sparse regularization. It is shown in our previous work th...

In recent years, deeper and wider neural networks have shown excellent performance in computer vision tasks, while their enormous amount of parameters results in increased computational cost and overfitting. Several methods have been proposed to compress the size of the networks without reducing network performance. Network pruning can reduce redun...

Lung cancer is one of the most deadly diseases in the world. Detecting such tumors at an early stage can be a tedious task. Existing deep learning architecture for lung nodule identification used complex architecture with large number of parameters. This study developed a cascaded architecture which can accurately segment and classify the benign or...

The detection of retinal blood vessels, especially the changes of small vessel condition is the most important indicator to identify the vascular network of the human body. Existing techniques focused mainly on shape of the large vessels, which is not appropriate for the disconnected small and isolated vessels. Paying attention to the low contrast...

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

The task of image annotation is becoming enormously important for efficient image retrieval from the web and other large databases. However, huge semantic information and complex dependency of labels on an image make the task challenging. Hence determining the semantic similarity between multiple labels on an image is useful to understand any incom...

The accurate detection of radioactive iodine-avid lymph node (LN) metastasis on 131I post-ablation whole-body planar scans (RxWBSs) is important in tracking the progression of the metastatic lymph nodes (mLNs) of patients with papillary thyroid cancer (PTC). However, severe noise artifacts and the indiscernible location of the mLN from adjacent tis...

Action recognition in videos is becoming popular these years. The difficulty is how to extract the temporal information, which is important in the target actions. In this paper, we propose a conceptually, simple network for short-term action recognition. The proposed network architecture is extended from standard neural network to Autoencoder, whic...

Convolutional neural network (CNN) have been extensively applied for a variety of tasks. However, the internal processes of hidden units in solving problems are mostly unknown. In this study, we presented the use of canonical correlation analysis (CCA) to understand the information flow of the hidden layers in CNN. The proposed method analyzed and...

Lung cancer is one of the most deadly diseases in the world. Detecting such tumors at an early stage can be a tedious task. Existing deep learning architecture for lung nodule identification used complex architecture with large number of parameters. This study developed a cascaded architecture which can accurately segment and classify the benign or...

A deep convolutional neural network (CNN) has been widely used in image classification and gives better classification accuracy than the other techniques. The softmax cross-entropy loss function is often used for classification tasks. There are some works to introduce the additional terms in the objective function for training to make the features...

In recent years, deep learning has spread rapidly, and deeper, larger models have been proposed. However, the calculation cost becomes enormous as the size of the models becomes larger. Various techniques for compressing the size of the models have been proposed to improve performance while reducing computational costs. One of the methods to compre...

In a deep neural network (DNN), the number of the parameters is usually huge to get high learning performances. For that reason, it costs a lot of memory and substantial computational resources, and also causes overfitting. It is known that some parameters are redundant and can be removed from the network without decreasing performance. Many sparse...

Deep convolutional neural network (CNN) with considerable number of parameters is one of the promising methods for image recognition. There, however, is generally difficult in applying deep CNNs to resource constrained devices due to the heavy computational burden. For reducing computational cost of CNNs while retaining the classification performan...

Anomaly detection is important to significant real life entities such as network intrusion and credit card fraud. Existing anomaly detection methods were partially learned the features, which is not appropriate for accurate detection of anomalies. In this study we proposed vector-based convolutional autoencoder (V-CAE) for one dimensional anomaly d...

Lung cancer is one of the most deadly diseases in the world. Detecting such tumors at an early stage can be a tedious task. Existing deep learning architecture for lung nodule identification used complex architecture with a large number of parameters. This study developed a cascaded architecture which can accurately segment and classify the benign...

Convolutional neural network (CNN) have been extensivelyapplied for a variety of tasks. However, the internal processes of hiddenunits in solving problems are mostly unknown. In this study, we pre-sented the use of canonical correlation analysis (CCA) to understandthe information flow of the hidden layers in CNN. The proposed methodanalyzed and com...

We propose an easy-to-use non-overlapping camera calibration method. First, successive images are fed to a PoseNet-based network to obtain ego-motion of cameras between frames. Next, the pose between cameras are estimated. Instead of using a batch method, we propose an on-line method of the inter-camera pose estimation. Furthermore, we implement th...

Assessing the structure and function of organelles in living organisms of the primitive unicellular red algae Cyanidioschyzon merolae on three-dimensional sequential images demands a reliable automated technique in the class imbalance among various cellular structures during mitosis. Existing classification networks with commonly used loss function...

This paper describes a network that captures multimodal correlations over arbitrary timestamps. The proposed scheme operates as a complementary, extended network over a multimodal convolutional neural network (CNN). Spatial and temporal streams are required for action recognition by a deep CNN, but overfitting reduction and fusing these two streams...

Multi-label image annotation based on convolutional neural networks (CNN) has seen significant improvements in recent years. One problem, however, is that it is difficult to prepare complete labels for the training images and usually training data has missing or incomplete labels. Restricted Boltzmann Machines (RBM) can explore the co-occurrence di...

It is known that RankSVM can optimize area under the ROC curve (AUC) for binary classification by maximizing the margin between the positive class and the negative class. Since the objective function of Siamese Network for rank learning is the same as RankSVM, Siamese Network can also optimize AUC for binary classification. This paper proposes a me...

The detection of retinal blood vessels, especially the changes of small vessel condition is the most important indicator to identify the vascular network of the human body. Existing techniques focused mainly on shape of the large vessels, which is not appropriate for the disconnected small and isolated vessels. Paying attention to the low contrast...

In recent years, number of classification techniques for Alzheimer's disease (AD) have been developed that produced methods based on the use of hand-crafted machine learning and obscure deep learning models. This study proposed a new classification framework based on the combination of Unet-like 2D convolutional neural networks (CNN) and multinomia...

Two-Dimensional Neural Network (2D CNN) has become an alternative method for one-dimensional data classification. Previous studies are focused either only on sequence or vector data. In this paper, we proposed a new 2D CNN classification method that suitable for both, sequence and vector data. The Hilbert space-filling curve was used as a 1D to 2D...

Segmentation of three-dimensional (3D) medical images using deep learning is a challenging task due to the lack of a 3D medical image dataset and their ground truth, resource memory limitations, and imbalanced dataset problem. In this paper, we propose advanced deep learning network for segmentation of 3D medical images. The proposed Multi-projecti...

Video Super resolution algorithms usually utilize the motion information of each pixel in consecutive frames to interpolate pixel values in a higher resolution and reconstruct frames of a higher resolution video from a lower resolution input video. Recently, architectures based on deep neural networks have gained popularity and can generate higher...