In this paper, we propose an acceleration-based human activity recognition method using popular deep architecture, Convolution Neural Network (CNN). particular, construct a CNN model and modify the convolution kernel to adapt characteristics of tri-axial acceleration signals. Also, for comparison, use some widely used methods accomplish task on same dataset. The large dataset constructed consists 31688 samples from eight typical activities. experiment results show that works well, which can...

Metal halides are widely applied in solid-state lighting (SSL), optoelectronic devices, information encryption, and near-infrared (NIR) detection due to their superior photoelectric properties tunable emission. However, single-component phosphors that can be efficiently excited by light-emitting diode (LED) chips cover both the visible (VIS) NIR emission regions still very rare. To address this issue, (TPA)2ZnBr4:Sn2+/Mn2+ (TPA = [(CH3CH2CH2)4N]+) were synthesized using solvent evaporation...

With the rapid development of mobile devices and pervasive computing technologies, acceleration-based human activity recognition, a difficult yet essential problem in apps, has received intensive attention recently. Different acceleration signals for representing different activities or even same have attributes, which causes troubles normalizing signals. We thus cannot directly compare these with each other, because they are embedded nonmetric space. Therefore, we present scheme that...

As an active research field, sport-related activity monitoring plays important role in people’s lives and health. This is often viewed as a human recognition task which fixed-length sliding window used to segment long-term signals. However, activities with complex motion states non-periodicity can be better monitored if the algorithm able accurately detect duration of meaningful states. this ability lacking approach. In study, we focused on two types for monitoring, regard detection task....

In this paper, a naturalistic 3D acceleration-based activity dataset, the SCUT-NAA is created to assist researchers in field of recognition and provide standard dataset for comparing evaluating performance different algorithms. The first publicly available contains 1278 samples from 44 subjects (34 males 10 females) collected settings with only one tri-axial accelerometer located alternatively on waist belt, trousers pocket, shirt pocket. Each subject was asked perform ten activities....

Sulfur-based autotrophic denitrification is a novel biological process characterized by the absence of an organic carbon source, short reaction time, high rate, low treatment cost, and small footprint. However, technique facing challenges with respect to engineering applications. In this study, pilot-scale sulfur-based system was established optimal hydraulic retention time (HRT) 0.21 h, which achieved highest load 1158 mg/(L·d) rate 164 gNO3−-N/(m3·h). Effective backwashing basis for...

In this paper, we provide two databases DB1 and DB2 of motion characters written in the air, present an accelerometers gyroscopes based air-writing recognition system. The 10 was collected by 40 subjects without writing constraints, while 36 49 participants constrained stroke orders. We preprocessed raw data with Moving Average filter Z-score normalization, then utilized a modified CHMM for modeling Viterbi algorithm recognition. ASD rule states assignment managed to avoid underflow issue...

Most existing researches on inertial sensor based dynamic motion gestures use deterministic or stochastic methods, however, these models generally possess short term memory so that they only memorize few time steps before and ignore the historical information deeper in time. Furthermore, researchers mainly investigate primary level gestures, while with higher complexity are more powerful expression. In this paper, we implement an end-to-end framework for recognition multi-complexity using a...

An algorithm for the discrimination between human upstairs and downstairs using a tri-axial accelerometer is presented in this paper, which consists of vertical acceleration calibration, extraction two kinds features (Interquartile Range Wavelet Energy), effective feature subset selection with wrapper approach, SVM classification. The proposed can recognize 95.64% average accuracy different sensor locations, i.e. located on subject's waist belt, trousers pocket, shirt pocket. Even mixed data...

Many single model methods have been applied to real-time short-term traffic flow prediction. However, since data is mixed with a variety of ingredients, the performance limited. Therefore, we proposed Multi-Long-Short Term Memory Models, which improved prediction accuracy comparing state-of-the-art models.

All tables can be represented as grids. Based on this observation, we propose GridFormer, a novel approach for interpreting unconstrained table structures by predicting the vertex and edge of grid. First, flexible representation in form an M X N In representation, vertexes edges grid store localization adjacency information table. Then, introduce DETR-style structure recognizer to efficiently predict multi-objective single shot. Specifically, given set learned row column queries, directly...

Clear cell renal carcinoma (ccRCC) originates from tubular epithelial cells and is the most common pathological type with worst prognosis. The relationship between expression, prognosis mechanism of ccRCC E2F family remains challenging. In present study, RNA sequencing clinical data Cancer Genome Atlas two datasets, GSE36895 GSE53757, Gene Expression Omnibus were used to identify role in ccRCC. A total 10 groups tumor tissues paired-normal patients verified by reverse...

To the best of our knowledge, there are a few researches on air-handwriting character-level writer identification only employing acceleration and angular velocity data. In this paper, we propose deep learning approach to using inertial sensor data air-handwriting. particular, separate different representations degree freedom (DoF) extract local dependency interrelationship in CNNs separately. Experiments public dataset achieve an average good performance without any extra hand-designed...