Most previous EEG-based emotion recognition methods studied hand-crafted EEG features extracted from different electrodes. In this article, we study the relation among electrodes and propose a deep learning method to automatically extract spatial that characterize functional between signals at Our proposed model is called <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AT</b> tention-based xmlns:xlink="http://www.w3.org/1999/xlink">LSTM</b>...

With the development of virtual reality (VR) and human-computer interaction technology, how to use natural efficient methods in environment has become a hot topic research. Gesture is one most important communication human beings, which can effectively express users' demands. In past few decades, gesture-based made significant progress. This article focuses on gesture technology discusses definition classification gestures, input devices for interaction, recognition technology. The...

In this paper a novel approach is described for tree visualization using nested circles. The brother nodes at the same level are represented by externally tangent circles; different levels displayed 2D circles or 3D cylinders. A new layout algorithm structure described. It provides good overview large data sets. easy to see all branches and leaves of tree. method has been applied file systems.

Hand detection is essential for many hand related tasks, e.g., recovering pose and understanding gesture. However, in uncontrolled environments challenging due to the flexibility of wrist joint cluttered background. We propose a convolutional neural network (CNN), which formulates in-plane rotation explicitly solve estimation jointly. Our architecture adopts backbone faster R-CNN generate rectangular region proposals extract local features. The takes feature as input estimates an manages...

In this paper, we propose a novel deep learning framework to reconstruct 3D hand poses and shapes of two interacting hands from single color image. Previous methods designed for cannot be easily applied the scenario because heavy inter-hand occlusion larger solution space. order address similar appearance between that may confuse network, design pose-aware attention module extract features associated each individual respectively. We then leverage context presented in interaction...

With the fast economic growth and urbanization of many developing countries come concerns that their children now have fewer opportunities to express creativity develop collaboration skills, or experience local cultural heritage. We propose address these by creating technologies inspired traditional arts, allowing create collaborate through playing with them. ShadowStory is our first attempt in this direction, a digital storytelling system Chinese shadow puppetry. present design...

We propose a novel 3D neural network architecture for hand pose estimation from single depth image. Different previous works that mostly run on 2D image domain and require intermediate or post process to bring in the supervision space, we convert map volumetric representation, feed it into convolutional network(CNN) directly produce requiring no further process. Our system does not ground truth reference point initialization, our naturally integrates both local feature global context space....

Hand-drawn sketch recognition is a fundamental problem in computer vision, widely used sketch-based image and video retrieval, editing, reorganization. Previous methods often assume that complete as input; however, hand-drawn sketches common application scenarios are incomplete, which makes challenging problem. In this paper, we propose SketchGAN, new generative adversarial network (GAN) based approach jointly completes recognizes sketch, boosting the performance of both tasks. Specifically,...

Chatbots systems, despite their popularity in today's HCI and CSCW research, fall short for one of the two reasons: 1) many systems use a rule-based dialog flow, thus they can only respond to limited number pre-defined inputs with pre-scripted responses; or 2) are designed focus on single-user scenarios, it is unclear how these may affect other users community. In this paper, we develop generalizable chatbot architecture (CASS) provide social support community members an online health The...

The compositions of the microbial community in fermented grains Sauce-flavor baijiu produced different regions have diverse characteristics; however, reasons for this remain unclear. present study investigated contributions environmental microorganisms to as well volatile compounds Beijing region using high-throughput sequencing combined with sourcetracker analysis, and compared differences microorganism their roles production process from regions.The results showed that tools were main...

Anomaly detection is currently an important and active research problem in many fields involved numerous applications. Most of the existing methods are based on distance measure. But case data stream these not very efficient as computational point view. exiting work outlier declare a outlier/inlier soon it arrive due to limited memory resources compared huge stream, often can lead us wrong decision, because dynamic nature incoming data. In this paper we introduced clustering approach, which...

Thermal control is crucial to real-time systems as excessive processor temperature can cause system failure or unacceptable performance degradation due hardware throttling. Real-time face significant challenges in thermal management they must avoid overheating while still delivering desired performance. Furthermore, many handle a broad range of uncertainties and environmental conditions. To address these challenges, this paper presents Control under Utilization Bound (TCUB), novel algorithm...

Smoothed particle hydrodynamics (SPH) is efficient, mass preserving, and flexible in handling topological changes. However, sparsely sampled thin features are difficult to simulate SPH-based free surface flows, due a number of robustness stability issues. In this article, we address problem from two perspectives: the forces numerical instability features. We present new tension force scheme based on energy functional, under diffuse interface model. develop an efficient way calculate air...

In this paper, we propose a Geometry-Contrastive Generative Adversarial Network (GC-GAN) for transferring continuous emotions across different subjects. Given an input face with certain emotion and target facial expression from another subject, GC-GAN can generate identity-preserving the expression. Geometry information is introduced into cGANs as conditions to guide generation of expressions. order handle misalignment subjects or emotions, contrastive learning used transform geometry...

3D hand pose estimation from single depth is a fundamental problem in computer vision, and has wide applications. However, the existing methods still can not achieve satisfactory results due to view variation occlusion of human hand. In this paper, we propose new virtual selection fusion module for depth. We automatically select multiple viewpoints fuse all find empirically delivers accurate robust estimation. order most effective views fusion, evaluate based on confidence using light-weight...

Daqu is a microbial-rich baijiu fermentation starter. The high-temperature plays an essential role in the manufacturing of sauce-flavored baijiu. However, few studies have focused on three kinds (white, yellow, and black Daqu) northern China. In this study, physicochemical indexes, volatile flavor compounds, microbial characteristics different colors China were comparatively analyzed to reveal their potential functions. White (WQ) exhibited highest liquefying power starch, (BQ) showed...

We present a new technique called 'Tilt Menu' for better extending selection capabilities of pen-based interfaces. The Tilt Menu is implemented by using 3D orientation information pen devices while performing tasks. has the potential to aid traditional one-handed techniques as it simultaneously generates secondary input (e.g., command or parameter selection) drawing/interacting with tip without having use second hand another device. conduct two experiments explore performance Menu. In first...

This paper presents a tangible programming tool 'T-Maze' for children aged 5 to 9. Children could use T-Maze create their own maze maps and complete some escaping tasks by the blocks sensors. uses camera to, in real-time, catch sequence of wooden blocks' arrangement, which will be used analyze semantic correctness enable receive feedbacks immediately. And join game controlling sensors during program's running. A user study shows that is an interesting approach easy learn use.

Turning over fermented grains is widely used during the heap fermentation of sauce-flavor baijiu to increase temperature, however, effect turning on still unclear. The present study investigated changes microbial community structure and volatile compounds before after by using high-throughput sequencing combined with gas chromatography-mass spectrometry (GC-MS). Firstly, lead a redistribution acidity moisture content in different layers an temperature. Secondly, total 41 were identified,...

3D hand pose estimation is a challenging problem in computer vision due to the high degrees-of-freedom of articulated motion space and large viewpoint variation. As consequence, similar poses observed from multiple views can be dramatically different. In order deal with this issue, view-independent features are required achieve state-of-the-art performance. paper, we investigate impact on single depth image, propose novel recurrent neural network for estimation, which cascaded pose-guided...

The inadequate use of finger properties has limited the input space touch interaction. By leveraging category contacting fingers, finger-specific interaction is able to expand vocabulary. However, accurate identification remains challenging, as it requires either additional sensors or sets identifiable fingers achieve ideal accuracy in previous works. We introduce SpeciFingers, a novel approach identify with capacitive raw data on touchscreens. apply neural network an encoder-decoder...