Deforestation rates in Brazilian Amazon continue to increase and addressing deforestation requires coordinated action at multiple levels, as well as international cooperation despite various strategies and action at multi-levels. To promote such strategies and action, this paper proposed a recurrent deep learning-based model, which predicts monthly deforestation events (will occur or not) at 1 km × 1 km meshes. The model was trained and evaluated by using about seven-year deforestation data (August 2016–September 2023) of eight areas in the Brazilian Amazon from the Real-Time Deforestation Detection System. By using the most recent two-year data, it was clear that the monthly prediction has a reasonable performance for recall, but very low precision. Thus, future work in the monthly prediction of deforestation should focus on refining model accuracy by incorporating more diverse data sources and improving temporal resolution.

An intention refinement method for domain-data analysis using large language models is proposed. It employs a “domain-framing prompt” to relate a domain expert's interests with relevant domain data. The initial responses are followed by “probing prompts” that suggest possible follow-up questions to provide further insight. The suggestions aim to support the domain expert in recognizing alternative analysis steps that may be difficult to notice depending on their analysis skills. The main contribution of this study is the proposal of this human-centered process to support domain experts in interacting with an LLM effectively. An experimental implementation demonstrates the method's effectiveness in expanding user interests and generating exploratory questions to assist intention refinement.

Reducing inter-subject variability between new users and the measured source subjects, and effectively using the information of classification models trained by source subject data, is very important for human–machine interfaces. In this study, we propose a style transfer mapping (STM) and fine-tuning (FT) subject transfer framework using convolutional neural networks (CNNs). To evaluate the performance, we used two types of public surface electromyogram datasets named MyoDatasets and NinaPro database 5. Our proposed framework, STM-FT-CNN, showed the best performances in all cases compared with conventional subject transfer frameworks. In the future, we will build an online processing system that includes this subject transfer framework and verify its performance in online experiments.

Surface electromyogram (sEMG)-based human-computer interface (HCI) is an effective tool for detecting human movements. Because sEMG-based motion recognition usually requires prolonged data measurements from the user (target), transfer learning reusing pre-measured (source) data from other users and pre-trained classifiers can be applied to sEMG data to reduce the measurement time. However, little knowledge is available regarding the combination of transfer learning methods and classifiers in sEMG data applications. Thus, we investigated the classification accuracy of data- and parameter-space-based transfer learning with shallow or deep classifiers in cross-subject sEMG classification. The dataset contains eight classes of forearm motions recorded from 25 volunteer participants. We used a support vector machine (SVM) as a shallow classifier as well as a deep neural network architecture, referred to as an artificial neural network (ANN), as a deep classifier. In addition, we used style transfer mapping (STM) as a data-space-based transfer learning method and fine-tuning (FT) as a parameter-space-based transfer learning method. Consequently, the classification accuracy of the ANN was higher than that of the SVM, regardless of the combinational use of transfer learning. STM and FT significantly improved the classification accuracy compared with non-transfer cases regardless of the classifier (note that FT can only be used with the ANN). In particular, the combined use of FT and the ANN yielded the best accuracy. These findings suggest that parameter-space-based transfer learning and deep classifiers are suitable for cross-subject sEMG classification. The combined use of parameter-space-based transfer learning and deep classifiers can effectively reduce the data measurement time of sEMG-based HCI applications.

To construct subject-specific feature extractors and classifiers for a new subject using pooled datasets, overcoming intersubject variabilities is required. In this study, we investigate the efficiency of the proposed subject transfer framework, which applies a discriminability-based source selection approach and semi-supervised style transfer mapping algorithm, by constructing support vector machine classifiers. We collect a surface electromyogram (sEMG) dataset acquired from 25 subjects using a wearable sEMG sensor. Classifiers are trained with gold-standard time-domain and autoregressive features extracted from eight-channel sEMG data. Compared with conventional subject transfer framework (85.08±1.38%), which applies the covariate shift adaptation algorithm to the linear discriminant analysis classifier and uses all source data, our proposed framework improves pattern recognition accuracy (90.63 ± 1.27%) by selection of discriminative source data and the mapping destination in the Euclidean space.

Functional near-infrared spectroscopy (fNIRS) offers a convenient means to measure human brain activity in real environments, driving many applications in neuroscience, medicine, and engineering. However, it is well known that fNIRS observations are always contaminated by scalp blood flow (SBF), which is not from the brain but from the skin. Several methods have been proposed for extracting cerebral blood flow (CBF), which reflects brain activities. The issue we tackled in this paper is how to evaluate the quality of CBF extraction for designing effective fNIRS applications. The standard way is to simultaneously measure reference signals such as SBF by laser Doppler blood flow meters and compare extracted CBF with them. This hardware approach requires additional instruments to be installed on the fNIRS device, declining the convenience of fNIRS recording in real environments. Then, in this paper we proposed a no-reference metric for evaluating the performance of CBF extraction. The proposed metric is based on decoding analysis of only fNIRS data and does not require additional instruments. The fundamental idea is to utilize the difference of SBF patterns in two experimental conditions; therefore an inherent limitation of the proposed metric is that it does not directly apply to single-condition data. We tested the proposed metric with popular CBF extraction techniques, namely, principal component analysis (PCA) and independent component analysis (ICA), and observed that these techniques may have room for improvement. The no-reference metric would be useful for evaluating CBF extraction techniques and promote the development of fNIRS applications.

In this study, we present a new fixture (self-attachable fixture) and tensile test system for measuring mechanical properties of cell sheet. To evaluate strength of cell sheet, it is the most important to measure mechanical properties of tensile mode. However, there has been no study which measured the tensile mechanical properties of cell sheet, since it has been difficult to attach a cell sheet in the tensile test system owing to the structure of the conventional fixture, and there has been no tensile test system which had a measurement range that covered the tension force range of the cell sheets. Therefore, we have addressed these problems by developing a self-attachable fixture and a tensile test system. By using developed system, we measured mechanical properties (tension, stress and initial stiffness) C2C12 of cell sheet cultured in different recipe of culture medium. The initial stiffness of cell sheet cultured in culture medium without FBS had a tendency to become stiffer. This indicates that our new fixture and test system are applicable for evaluating mechanical properties of cell sheets.

Here we present an environmentally robust mechano-bionic system using insect dorsal vessel tissues. Although mechano-bionic systems using mammalian heart muscle cells were reported previously, these mechano-bionic systems require precise environmental control to keep the heart muscle cells contracting spontaneously. To overcome this problem, our group focused on insect dorsal vessel tissues as a novel driving source of mechano-bionic system. We have shown so far that insect dorsal vessel can be utilized as driving source that is robust over culture conditions. In this paper, as a mechano-bionic system that is robust over its operating environment, we propose a micro-actuated system that consists of insect dorsal vessel and mechanical structure fabricated by micro fabrication technology. By photolithographic technique, the components of the machine were fabricated. The excised dorsal vessel was attached to the microstructure by tweezers under the stereomicroscope. The micro-mechanical components with dorsal vessel was incubated at 25°C, and after several hours of culture, it was actuated by contraction of dorsal vessel. Using image analysis, its maximum open-end displacement and operating frequency was measured around 27μm and 0.5Hz, respectively. We successfully demonstrated in constructing the mechano-bionic system that is robust over its operating environment.

A three-dimensional cardiomyocyte gel was reconstructed using an extracellular matrix gel formed from Matrigel and collagen I. The autonomous beating of the cardiomyocyte gel was synchronized with contraction of cardiomyocytes in the gel. The gel beating frequency was increased by increasing the concentration of adrenaline in the cardiomyocyte culturing medium. Drug stimulation to modulate the beating frequency could be used for control of a pumping application. A micropump driven by the cardiomyocyte gel was demonstrated. The micropump consisted of two check valves and the cardiomyocyte gel actuator. The volume changes of the channel in the micropump, synchronized with the autonomous gel beating, were seen.