Multi-Spectral Food Classification and Caloric Estimation Using Predicted Images.

In nutrition science, methods that accomplish continuous recognition of ingested foods with minimal user intervention have great utility. Our recent study showed that using images taken at a variety of wavelengths, including ultraviolet (UV) and near-infrared (NIR) bands, improves the accuracy of food classification and caloric estimation. With this approach, however, analysis time increases as the number of wavelengths increases, and there are practical implementation issues associated with a large number of light sources. To alleviate these problems, we proposed a method that used only standard red-green-blue (RGB) images to achieve performance that approximates the use of multi-wavelength images. This method used RGB images to predict the images at each wavelength (including UV and NIR bands), instead of using the images actually acquired with a camera. Deep neural networks (DNN) were used to predict the images at each wavelength from the RGB images. To validate the effectiveness of the proposed method, feasibility tests were carried out on 101 foods. The experimental results showed maximum recognition rates of 99.45 and 98.24% using the actual and predicted images, respectively. Those rates were significantly higher than using only the RGB images, which returned a recognition rate of only 86.3%. For caloric estimation, the minimum values for mean absolute percentage error (MAPE) were 11.67 and 12.13 when using the actual and predicted images, respectively. These results confirmed that the use of RGB images alone achieves performance that is similar to multi-wavelength imaging techniques.

Multispectral Food Classification and Caloric Estimation Using Convolutional Neural Networks.

Continuous monitoring and recording of the type and caloric content of ingested foods with a minimum of user intervention is very useful in preventing metabolic diseases and obesity. In this paper, automatic recognition of food type and caloric content was achieved via the use of multi-spectral images. A method of fusing the RGB image and the images captured at ultra violet, visible, and near-infrared regions at center wavelengths of 385, 405, 430, 470, 490, 510, 560, 590, 625, 645, 660, 810, 850, 870, 890, 910, 950, 970, and 1020 nm was adopted to improve the accuracy. A convolutional neural network (CNN) was adopted to classify food items and estimate the caloric amounts. The CNN was trained using 10,909 images acquired from 101 types. The objective functions including classification accuracy and mean absolute percentage error (MAPE) were investigated according to wavelength numbers. The optimal combinations of wavelengths (including/excluding the RGB image) were determined by using a piecewise selection method. Validation tests were carried out on 3636 images of the food types that were used in training the CNN. As a result of the experiments, the accuracy of food classification was increased from 88.9 to 97.1% and MAPEs were decreased from 41.97 to 18.97 even when one kind of NIR image was added to the RGB image. The highest accuracy for food type classification was 99.81% when using 19 images and the lowest MAPE for caloric content was 10.56 when using 14 images. These results demonstrated that the use of the images captured at various wavelengths in the UV and NIR bands was very helpful for improving the accuracy of food classification and caloric estimation.

Improved spin-orbit torque induced magnetization switching efficiency by helium ion irradiation.

Increasing the efficiency of spin-orbit torque (SOT) is of great interest in spintronics devices because of its application to the non-volatile magnetic random access memory and in-logic memory devices. Accordingly, there are several studies to alter the magnetic properties and reduce the SOT switching current with helium ion irradiation, but previous researches are focused on its phenomenological changes only. Here, the authors observe the reduction of switching current and analyze its origins. The analyzed major reasons are improved spin Hall angle represented as the changed resistivity of heavy metal layer and the reduction of surface anisotropy energy at interface between heavy metal and ferromagnet. It is confirmed that almost linear relation between changed SHA and Pt resistivity by helium ion irradiation, which is attributed because of the increase in the scattering sources induced by structural distortion during ion penetration. From the calculated power consumption ratio based on the derived parameter, the requiring power decreases according to the degree of ion irradiation. Our results show that helium ion penetration induced layer and interfacial disturbance affects SOT induced magnetization switching current reduction and may provide possibility about helium ion irradiation based superior SOT device engineering.

Joint Audio-Ultrasound Food Recognition for Noisy Environments.

Continuous recognition of ingested foods without user intervention is very useful for the pre-screening of obesity and diet-related disease. An automatic food recognition method that combines the two modalities of audio and ultrasonic signals (US) is proposed in this study. Under a noise-free environment, classification accuracy of an audio-only recognizer is generally higher than that of US-only recognizers, but the performance of US recognizers is unaffected by acoustic noise levels. In the recognition system presented herein, the likelihood score of the audio-US feature was given by a linear combination of class-conditional observation log-likelihoods for two classifiers, using the appropriate weights. We developed a weighting process adaptive to signal-to-noise ratios (SNRs). The main objective here involves determining the optimal SNR classification boundaries and constructing a set of optimum stream weights for each SNR class. A feasibility test was conducted to verify the usefulness of the proposed method by conducting recognition experiments on seven types of food. The performance was compared with conventional methods that use in-ear and throat microphones. The proposed method yielded remarkable levels of recognition performance of 90.13% for artificially added noise and 89.67% under actual noisy environments, when the SNR ranged from 0 to 20 dB.

Field-free switching of perpendicular magnetization through spin-orbit torque in antiferromagnet/ferromagnet/oxide structures.

Spin-orbit torques arising from the spin-orbit coupling of non-magnetic heavy metals allow electrical switching of perpendicular magnetization. However, the switching is not purely electrical in laterally homogeneous structures. An extra in-plane magnetic field is indeed required to achieve deterministic switching, and this is detrimental for device applications. On the other hand, if antiferromagnets can generate spin-orbit torques, they may enable all-electrical deterministic switching because the desired magnetic field may be replaced by their exchange bias. Here we report sizeable spin-orbit torques in IrMn/CoFeB/MgO structures. The antiferromagnetic IrMn layer also supplies an in-plane exchange bias field, which enables all-electrical deterministic switching of perpendicular magnetization without any assistance from an external magnetic field. Together with sizeable spin-orbit torques, these features make antiferromagnets a promising candidate for future spintronic devices. We also show that the signs of the spin-orbit torques in various IrMn-based structures cannot be explained by existing theories and thus significant theoretical progress is required.

Angular and temperature dependence of current induced spin-orbit effective fields in Ta/CoFeB/MgO nanowires.

Current induced spin-orbit effective magnetic fields in metal/ferromagnet/oxide trilayers provide a new way to manipulate the magnetization, which is an alternative to the conventional current induced spin transfer torque arising from noncollinear magnetization. Ta/CoFeB/MgO structures are expected to be useful for non-volatile memories and logic devices due to its perpendicular anisotropy and large current induced spin-orbit effective fields. However many aspects such as the angular and temperature dependent phenomena of the effective fields are little understood. Here, we evaluate the angular and temperature dependence of the current-induced spin-orbit effective fields considering contributions from both the anomalous and planar Hall effects. The longitudinal and transverse components of effective fields are found to have strong angular dependence on the magnetization direction at 300 K. The transverse field decreases significantly with decreasing temperature, whereas the longitudinal field shows weaker temperature dependence. Our results reveal important features and provide an opportunity for a more comprehensive understanding of current induced spin-orbit effective fields.

Prediction of acoustic feature parameters using myoelectric signals.

It is well-known that a clear relationship exists between human voices and myoelectric signals (MESs) from the area of the speaker's mouth. In this study, we utilized this information to implement a speech synthesis scheme in which MES alone was used to predict the parameters characterizing the vocal-tract transfer function of specific speech signals. Several feature parameters derived from MES were investigated to find the optimal feature for maximization of the mutual information between the acoustic and the MES features. After the optimal feature was determined, an estimation rule for the acoustic parameters was proposed, based on a minimum mean square error (MMSE) criterion. In a preliminary study, 60 isolated words were used for both objective and subjective evaluations. The results showed that the average Euclidean distance between the original and predicted acoustic parameters was reduced by about 30% compared with the average Euclidean distance of the original parameters. The intelligibility of the synthesized speech signals using the predicted features was also evaluated. A word-level identification ratio of 65.5% and a syllable-level identification ratio of 73% were obtained through a listening test.

SNR-adaptive stream weighting for audio-MES ASR.

Myoelectric signals (MESs) from the speaker's mouth region have been successfully shown to improve the noise robustness of automatic speech recognizers (ASRs), thus promising to extend their usability in implementing noise-robust ASR. In the recognition system presented herein, extracted audio and facial MES features were integrated by a decision fusion method, where the likelihood score of the audio-MES observation vector was given by a linear combination of class-conditional observation log-likelihoods of two classifiers, using appropriate weights. We developed a weighting process adaptive to SNRs. The main objective of the paper involves determining the optimal SNR classification boundaries and constructing a set of optimum stream weights for each SNR class. These two parameters were determined by a method based on a maximum mutual information criterion. Acoustic and facial MES data were collected from five subjects, using a 60-word vocabulary. Four types of acoustic noise including babble, car, aircraft, and white noise were acoustically added to clean speech signals with SNR ranging from -14 to 31 dB. The classification accuracy of the audio ASR was as low as 25.5%. Whereas, the classification accuracy of the MES ASR was 85.2%. The classification accuracy could be further improved by employing the proposed audio-MES weighting method, which was as high as 89.4% in the case of babble noise. A similar result was also found for the other types of noise.

EMG-based speech recognition using hidden markov models with global control variables.

It is well known that a strong relationship exists between human voices and the movement of articulatory facial muscles. In this paper, we utilize this knowledge to implement an automatic speech recognition scheme which uses solely surface electromyogram (EMG) signals. The sequence of EMG signals for each word is modelled by a hidden Markov model (HMM) framework. The main objective of the work involves building a model for state observation density when multichannel observation sequences are given. The proposed model reflects the dependencies between each of the EMG signals, which are described by introducing a global control variable. We also develop an efficient model training method, based on a maximum likelihood criterion. In a preliminary study, 60 isolated words were used as recognition variables. EMG signals were acquired from three articulatory facial muscles. The findings indicate that such a system may have the capacity to recognize speech signals with an accuracy of up to 87.07%, which is superior to the independent probabilistic model.