Search

We present a novel gated recurrent neural network to detect when a person is chewing on food. We implemented the neural network as a custom analog integrated circuit in a 0.18 μm CMOS technology. The neural network was trained on 6.4 hours of data collected from a contact microphone that was mounted on volunteers' mastoid bones. When tested on 1.6 hours of previously-unseen data, the analog neural network identified chewing events at a 24-second time resolution. It achieved a recall of 91% and an F1-score of 94% while consuming 1.1 μW of power. A system for detecting whole eating episodes-like meals and snacks-that is based on the novel analog neural network consumes an estimated 18.8 μW of power.

The present study sought to understand the spatiotemporal characteristics, associated with changes in drought disasters during the Ming and Qing Dynasties in North China. The grade sequence of drought disasters at 21 sites for the given period (1470–1912 AD) in North China was studied herein. An ensemble empirical mode decomposition (EEMD) was used to analyze the multiple time-scales towards generating a simple and stable intrinsic modal function component. Comparisons and analysis of the frequency and intensity of drought disasters were made using polynomial fitting curves to understand the temporal variations of drought disasters. Two aspects were explored to study the spatial distribution and characteristics of drought disasters. The reconstruction of the sequence of drought disasters was based on the Empirical Orthogonal Function (EOF) and Rotated Empiric Orthogonal Function (REOF). The drought disaster was divided into several space modes and sensitive areas. Drought frequency was recurrent in the northern and low in the southern part of North China. Findings revealed drought frequency and intensity were high in southeast and low in the southwestern part of North China. The study would inform decisions on disaster prevention and mitigation thereby serving as a baseline print for predicting modern drought disasters.

Observed climate data are processed under the assumption that their time series are stationary, as in multi-step temperature and precipitation prediction, which usually leads to low prediction accuracy. If a climate system model is based on a single prediction model, the prediction results contain significant uncertainty. In order to overcome this drawback, this study uses a method that integrates ensemble prediction and a stepwise regression model based on a mean-valued generation function. In addition, it utilizes empirical mode decomposition (EMD), which is a new method of handling time series. First, a non-stationary time series is decomposed into a series of intrinsic mode functions (IMFs), which are stationary and multi-scale. Then, a different prediction model is constructed for each component of the IMF using numerical ensemble prediction combined with stepwise regression analysis. Finally, the results are fit to a linear regression model, and a short-term climate prediction system is established using the Visual Studio development platform. The model is validated using temperature data from February 1957 to 2005 from 88 weather stations in Guangxi, China. The results show that compared to single-model prediction methods, the EMD and ensemble prediction model is more effective for forecasting climate change and abrupt climate shifts when using historical data for multi-step prediction.

Poor eating habits in children and teenagers can lead to obesity, eating disorders, or life-threatening health problems. Although researchers have studied children’s eating behavior for decades, the research community has had limited technology to support the observation and measurement of fine-grained details of a child’s eating behavior. In this paper, we present the feasibility of adapting the Auracle, an existing research-grade earpiece designed to automatically and unobtrusively recognize eating behavior in adults, for measuring children’s eating behavior. We identified and addressed several challenges pertaining to monitoring eating behavior in children, paying particular attention to device fit and comfort. We also improved the accuracy and robustness of the eating-activity detection algorithms. We used this improved prototype in a lab study with a sample of 10 children for 60 total sessions and collected 22.3 hours of data in both meal and snack scenarios. Overall, we achieved an accuracy exceeding 85.0% and an F1 score exceeding 84.2% for eating detection with a 3-second resolution, and a 95.5% accuracy and a 95.7% F1 score for eating detection with a 1-minute resolution.

In this paper, we propose Auracle, a wearable earpiece that can automatically recognize eating behavior. More specifically, in free-living conditions, we can recognize when and for how long a person is eating. Using an off-the-shelf contact microphone placed behind the ear, Auracle captures the sound of a person chewing as it passes through the bone and tissue of the head. This audio data is then processed by a custom analog/digital circuit board. To ensure reliable (yet comfortable) contact between microphone and skin, all hardware components are incorporated into a 3D-printed behind-the-head framework. We collected field data with 14 participants for 32 hours in free-living conditions and additional eating data with 10 participants for 2 hours in a laboratory setting. We achieved accuracy exceeding 92.8% and F1 score exceeding 77.5% for eating detection. Moreover, Auracle successfully detected 20-24 eating episodes (depending on the metrics) out of 26 in free-living conditions. We demonstrate that our custom device could sense, process, and classify audio data in real time. Additionally, we estimate Auracle can last 28.1 hours with a 110 mAh battery while communicating its observations of eating behavior to a smartphone over Bluetooth.

The eastern Northwest Chinese drought/flood grade sequence was reconstructed according to historical data on drought and flood disasters for the period from 1470 to 1912. The changing characteristics of drought/flood disasters over 443 years in eastern Northwest China were analyzed by employing methods such as the moving average, accumulative anomaly, moving t test, and wavelet analysis. To verify the reliability of this sequence, long-term trends were analyzed and compared using precipitation sequences reconstructed for Huashan and the drought/flood index for Longxi. The drought/flood grade curve and stalagmite oxygen isotopic records of two provinces were also compared. The results show that: (1) drought disasters were severe and frequent in eastern Northwest China. An alternating pattern between drought and flood occurrences was observed, which can be divided into four stages: 1470–1540, 1541–1615, 1616–1796, and 1797–1912. We further classified the four stages into 12 partial drought or partial flood periods. (2) Twelve abrupt changes were observed on the decadal scale and five abrupt changes were detected on the half-century scale. (3) The reconstructed sequences display a periodic variability on multiple scales; notable variability was detected on the following timescales: 130–110 years, 70–80 years, 50–52 years, 28 years, 10–15 years, and 5 years. The first four primary periods were 28 years, 15 years, 73 years, and 5 years; 28 years shows the strongest oscillation. (4) Based on the comparison and analysis, the reconstructed drought/flood grade sequence is highly consistent with other precipitation sequences. It corresponds to stalagmite oxygen isotopes records of the same period. All these findings indirectly confirm the reliability of the recovered historical flood and drought grade sequence based on historical data.

Climate variability has become a hot topic worldwide in recent years. Although large numbers of climate data series have been reconstructed based on hundreds to thousands, or even tens of thousands, of years, our understanding of this problem is still controversial. We use the precipitation index (PI) series to study the periodicity of the precipitation in northern China from 1870 to 2002 and explore the climate variability on a large timescale. The analysis shows that the precipitation has periods of 2.27–3.03, 7.14, 10.00, 11.11, 12.50, 14.29, 16.67, 20.00, and 25.00 a. Based on complete ensemble empirical mode decomposition (CEEMD), the Pacific sea surface temperature (SST) and Pacific interdecadal oscillation (PDO) are decomposed into different frequencies. The results show that the SST and PDO have interannual to interdecadal periodicity. To determine the impact of the Pacific SST and PDO on the PI, we make use of the cross wavelet power spectrum and wavelet coherency spectrum to analyze their period relation and reveal their periodic change characteristics; it is found that different bands of the Pacific SST, PDO, and PI have high power.

This study analyses the temporal and spatial distributions of droughts and floods in northern China during the Qing Dynasty (1636–1912), with data ranging from 1644 to 1911, or nearly the entire dynastic period. Variations in the intensities of droughts and floods are investigated across different timescales and with respect to their spatial variation. The impacts and physical mechanisms of solar activity as well as surface temperatures of the Pacific Ocean on drought and flood intensities in northern China are analysed using a cross-wavelet method. The results reveal periodicities on scales of 10–11, 20–22, 45–50 and 90–100 yrs. Seven rapid changes are found on the 10-yr scale and four are found on the 30-yr scale. Analysing the spatial distribution of drought and flood intensities, we found a trend of decreasing precipitation from east to west. Precipitation over the Shandong region, which is in the eastern part of the study area, is significantly higher than that of the other regions, which may be explained by its coastal location and the resultant rich water vapour content of the atmosphere. Analyses of cross-wavelet power spectra and coherence spectra among the different drought and flood intensities and the Nino3 regional sea surface temperature (SST) index series and sunspot series indicate significant correlations between the drought and flood series and the Nino3 SST index series at scales of 2–4 and 8–12 yrs, and the phase difference differs across the different timescales. Furthermore, the drought and flood series and the sunspot series are closely correlated at a scale of 8–14 yrs.

Researchers strive to understand eating behavior as a means to develop diets and interventions that can help people achieve and maintain a healthy weight, recover from eating disorders, or manage their diet and nutrition for personal wellness. A major challenge for eating-behavior research is to understand when, where, what, and how people eat. In this paper, we evaluate sensors and algorithms designed to detect eating activities, more specifically, when people eat. We compare two popular methods for eating recognition (based on acoustic and electromyography (EMG) sensors) individually and combined. We built a data-acquisition system using two off-the-shelf sensors and conducted a study with 20 participants. Our preliminary results show that the system we implemented can detect eating with an accuracy exceeding 90.9% while the crunchiness level of food varies. We are developing a wearable system that can capture, process, and classify sensor data to detect eating in real-time.

Chronic disease is one of the most pressing health challenges facing the United States (and an increasing set of other countries). The onset or progression of diseases like obesity, diabetes, and metabolic disorder are strongly related to eating behavior, and scientists are still trying to fully understand the complex mixture of diet, exercise, genetics, sociocultural context, and physical environment that lead to these diseases. Health science, however, has no effective means for automatically measuring eating behavior in free-living conditions. The Auracle aims to be a wearable earpiece that detects eating behavior, to be fielded by health-science researchers in their efforts to study eating behavior and ultimately to develop interventions useful to individuals striving to address chronic disease related to eating.

This paper studies grade sequence of drought disasters in 21 sites in Ming and Qing Dynasties (1470–1912) in North China. Two aspects are explored, in order to study the spatial distribution and characteristics of drought disaster in North China. The reconstruction of the sequence of drought disaster in North China during the Ming and Qing Dynasties was based on Empirical Orthogonal Function (EOF) and on Rotated Empiric Orthogonal Function (REOF). The drought disaster has been divided into several space models and into several sensitive space areas. It can provide an important basis for the better understanding of the spatial distribution of drought disasters in North China during the historical period. The research's results show that: the frequency is high in northern area and is low in southern area of North China. The frequency of drought intensity is high in South-East, low in West China. Meanwhile, the North China can be divided into six main sensitive regions: Middle-east of North China, west of North China, South of North China, East of North China, North-east and North-west of North China.

The spatial distribution of automatic weather stations in regions of western China (e.g., Tibet and southern Xingjiang) is relatively sparse. Due to the considerable spatial variability of precipitation, estimations of rainfall that are interpolated in these areas exhibit considerable uncertainty based on the current observational networks. In this paper, a new statistical method for estimating precipitation is introduced that integrates satellite products and in situ observation data. This method calculates the differences between raster data and point data based on the theory of data assimilation. In regions in which the spatial distribution of automatic weather stations is sparse, a nonparametric kernel-smoothing method is adopted to process the discontinuous data through correction and spatial interpolation. A comparative analysis of the fusion method based on the double-smoothing algorithm proposed here indicated that the method performed better than those used in previous studies based on the average deviation, root mean square error, and correlation coefficient values. Our results indicate that the proposed method is more rational and effective in terms of both the efficiency coefficient and the spatial distribution of the deviations.