Noriji Kato

Poly-Silicon Thin-Film Transistors with Uniform Performance Fabricated by Excimer Laser Annealing

Uniform performance in poly-Si thin-film transistors (TFTs) has been successfully achieved by excimer laser annealing. Mobility and its uniformity over a substrate were 59±3 cm2/Vs for n-channel TFTs and 45±5 cm2/Vs for p-channel types. To achieve uniform performance, we combined step annealing that uses two energy levels, and small-pitch annealing that moves a beam forward by a small pitch. The proposed method can improve surface morphology and uniformity of grain size in poly-Si. A 400-stage CMOS shift register composed of these TFTs could operate at 5 V, and attained the speed of 1 MHz at 8 V.

Visual trigger templates for knowledge-based indexing

We present an application to create binary Visual Trigger Templates (VTT) for automatic video indexing. Our approach is based on the observation that videos captured with fixed cameras have specific structures that depend on world constraints. Our system allows a user to graphically represent such constraints to automatically recognize simple actions or events. VTTs are constructed by manually drawing rectangles to define trigger spaces: when elements (e.g., a hand, a face) move inside the trigger spaces defined by the user, actions are recognized. For example, a user can define a raise hand action by drawing two rectangles: one for the face and one for the hand. Our approach uses motion, skin, and face detection algorithms. We present experiments on the PETS-ICVS dataset and on our own dataset to demonstrate that our system constitutes a simple but powerful mechanism for meeting video indexing.

Gender estimation for SNS user profiling using automatic image annotation

User profiling for Social Network Services (SNS) has gained great attention because of its potential values in identifying target population, which is very informative for marketing. Many studies have been conducted to estimate SNS user profiles using text analysis. However, in spite of the huge quantities of image resources on SNS, no previous work has specifically explored user profiles by automatic image annotation techniques. This paper addresses the problem of inferring a SNS users gender by automatic image annotation. The proposed method involves learning a model to annotate SNS images and integrating annotation scores of images to infer a users gender. Evaluation based on Twitter data demonstrates promising results.

Bag-of-features approach for improvement of lung tissue classification in diffuse lung disease

Many automated techniques have been proposed to classify diffuse lung disease patterns. Most of the techniques utilize texture analysis approaches with second and higher order statistics, and show successful classification result among various lung tissue patterns. However, the approaches do not work well for the patterns with inhomogeneous texture distribution within a region of interest (ROI), such as reticular and honeycombing patterns, because the statistics can only capture averaged feature over the ROI. In this work, we have introduced the bag-of-features approach to overcome this difficulty. In the approach, texture images are represented as histograms or distributions of a few basic primitives, which are obtained by clustering local image features. The intensity descriptor and the Scale Invariant Feature Transformation (SIFT) descriptor are utilized to extract the local features, which have significant discriminatory power due to their specificity to a particular image class. In contrast, the drawback of the local features is lack of invariance under translation and rotation. We improved the invariance by sampling many local regions so that the distribution of the local features is unchanged. We evaluated the performance of our system in the classification task with 5 image classes (ground glass, reticular, honeycombing, emphysema, and normal) using 1109 ROIs from 211 patients. Our system achieved high classification accuracy of 92.8%, which is superior to that of the conventional system with the gray level co-occurrence matrix (GLCM) feature especially for inhomogeneous texture patterns.

A new method to estimate grain boundary trap state density in poly-Si TFTs

A new method of estimating grain boundary trap state density in poly-Si thin film transistors (TFTs) is proposed by modifying an assumption used in Levinsons method. Our method assumes that only the carrier near the surface contributes to the total current, and the other carrier is neglected. Then the carrier density at the surface is used to express the potential barrier height induced at the grain boundary instead of the averaged carrier density as in Levinsons method. The validity of our assumption is investigated using our two-dimensional device simulator, and it is shown that our assumption on the carrier density is suitable to derive the true trap state density.

Classification of human actions using face and hands detection

In this paper, we describe a novel classification technique that separates video scenes, like office work tasks, into several scenes according to each task. Even if the difference of as a whole image frame by frame in each task is small, the difference of workers movement is quite big due to the position of face and hands according to each task. In addition, the worker has the tendency to turn his/her face to look at the particular objects of individual tasks like PC, a document, and so on. Then, we decide to separate tasks based on face position, face angle in depth, and hand positions. For comparison of frames in a video, we use the Maharanobis distance to measure the difference of multivariate data that consist of face coordinates, face angle in depth, and coordinates of both hands. For the separation of tasks by the Maharanobis distance of an each frame, we use the hierarchical clustering method to classify frames in a video according to each task. For the robust detection of both hands, we use color-based method that searches hand areas using face color. Although the color of hands changes corresponding to the lighting conditions, the color of hands should be very similar to that of the face in an office. Therefore, even when the lighting condition changes, our color-based hand detection method does not need any adjustment to the change. We apply this classification technique to separate office work video into individual sets of task scenes. As a result, our technique shows better task separation performance than the histogram-based boundary detection technique.

An analysis-synthesis loop model using kernel method

An analysis-synthesis loop model is constructed in feature space into which an input vector is mapped by a particular non-linear function. In this model, the recognition process is realized by comparison of the mapped input vector to reconstructed vectors that are generated initially and refined iteratively by a linear combination of the basis vectors in feature space. The kernel method allows efficient computation of the analysis-synthesis loop in the high dimensional feature space. Some experiments based on our model show more effective recognition of real-world images than that based on the linear model.

A Fabrication of Homogenious Poly-Si TFT's Using Excimer Laser Annealing

The poly-Si TFT is expected to provide driving circuits for large-area devices such as image sensors and LC displays. Thoug h an excimer laser crystallized poly-Si TFT has high performancel), there have been few reports about its uniformity.An energy deviation within a beam causes variation of grain size in the polySi and the crystallinity degrades in the overlapped area between beams2). These would degrade uniformity of TFT performance. Moreover, after this crystallization, su rface morphology is poor and the roughness increases as the laser energy is increased. We have evaluated the dependence of uniformity on laser energy and beam scanning pitch. As a result , both step annealingl) that uses two energy levels, and small pitch annealingal that moves a beam forward by a small pitch, have been effective for uniformity . In this paper, to obtain excellent uniformity for CMOS circuits, we have combined step annealing and small pitch method, and optimized step annealing conditions with small pitch annealing .

Content-Aware Multi-task Neural Networks for User Gender Inference Based on Social Media Images

To estimate demographic attributes such as gender and age of social media users from images posted by the users is a challenging problem because the demographic attributes are directly not shown in images. For such problem, prior approaches can be roughly separated into two types: one approach uses concept detection to detect pre-defined visual concepts which are then used as meta-data to estimate demographic attributes and the other approach directly uses content features such as Fisher Vector [19] which are extracted from images. In this paper we consider the way of combining these two approaches. We propose Multi-task Bilinear Model for integrating the detected concepts with the content features. In our proposed method, both the concept detector and the feature extractor can be jointly learned with end-to-end fashion. We evaluated the proposed method for the task of estimating user gender from Twitter images and found that it outperformed other baseline methods.

Size-Independent Image Segmentation by Hierarchical Clustering and Its Application for Face Detection

In this paper, we introduce a technique to detect a target object quickly. Our idea is based on onservation on the clusters into which an image is divided by hierarchical k-means clustering with space feature and color feature. This clustering method has the advantage of extracting the region of an object with some varied size. We insist that our idea should lead to detect a target object quickly, because it is not necessary to search the locations containing no targets. First, we evaluate our clustering method and second, we demonstrate that our method is effective on an object detection by applying to our face detection system. We show that the detection time can be reduced by 24%.