Search Results (142)

To improve the accuracy of Japanese finger-spelled character recognition using an RGB camera, we focused on feature design and refinement of the recognition method. By leveraging angular features extracted via MediaPipe, we proposed a method that effectively captures subtle motion differences while minimizing the influence of background and surrounding individuals. We constructed a large-scale dataset that includes not only the basic 50 Japanese syllables but also those with diacritical marks, such as voiced sounds (e.g., “ga”, “za”, “da”) and semi-voiced sounds (e.g., “pa”, “pi”, “pu”), to enhance the model’s ability to recognize a wide variety of characters. In addition, the application of a change-point detection algorithm enabled accurate segmentation of sign language motion boundaries, improving word-level recognition performance. These efforts laid the foundation for a highly practical recognition system. However, several challenges remain, including the limited size and diversity of the dataset and the need for further improvements in segmentation accuracy. Future work will focus on enhancing the model’s generalizability by collecting more diverse data from a broader range of participants and incorporating segmentation methods that consider contextual information. Ultimately, the outcomes of this research should contribute to the development of educational support tools and sign language interpretation systems aimed at real-world applications. Full article

Understanding the activity patterns of a species is essential for developing sound conservation and management plans. In this study, we used a camera-trapping technique to determine the activity patterns of bharal (Pseudois nayaur) in a marginal population in Wolong National Nature Reserve, Sichuan, China. Our results showed that these animals preferred to be active in the daytime from 08:00 to 20:00, with an activity peak between 10:00 and 14:00. In addition, we found that the species had a seasonal activity pattern with higher activity frequency in summer than in winter and that bharal were most active in a temperature range of 3–11 °C and at night with a waxing crescent moon, implying that the activity rhythm of the species is an adaptation to a subtropical high-altitude alpine area with vertical zonation in temperature. The pattern of movement and activity was also correlated with the moon phase. Full article

Sound source localization represents one of the major challenges for soundscapes due to the dynamicity of a large variety of signals. Many applications are found related to ecosystems to study the migration process of birds and animals other than other terrestrial environments to survey wildlife. Other applications on sound recording are supported by sensors to detect animal movement. This paper deals with the immersive 3D soundscape by using a multi-channel spherical microphone probe, in combination with a 360° camera. The soundscape has been carried out in three Italian squares across the city of Parma. The acoustic maps obtained from the data processing detect the directivity of dynamic sound sources as typical of an urban environment. The analysis of the objective environmental parameters (like loudness, roughness, sharpness, and prominence) was conducted alongside the investigations on the historical importance of Italian squares as places for social inclusivity. A dedicated listening playback is provided by the AGORA project with a portable listening room characterized by modular unit of soundbars. Full article

The article discusses the causes of excessive noise and vibrations of a live steam pipeline in a power unit. A scanning laser vibrometer was used to measure the vibrations of the live steam pipeline for two power units. Additionally, the sound (noise) level of the live steam pipeline was measured with an acoustic camera. A discrete model of the pipeline was created, and FEM modal analysis was performed. Based on experimental tests and numerical simulations, the sources of noise were identified. The final conclusions propose methods of eliminating the harmful noise. Full article

As robotic systems become more prevalent in daily life and industrial environments, ensuring their reliability through autonomous self-diagnosis is becoming increasingly important. This study investigates whether acoustic sensing can serve as a viable foundation for such self-diagnostic systems by examining its effectiveness in localizing structural faults. This study focuses on developing a fault diagnosis framework for robots using acoustic sensing technology. The objective is to design a simple yet accurate system capable of identifying fault locations and types of robots based solely on sound data, without relying on traditional sensors or cameras. To achieve this, sweep signals were applied to a modular robot, and acoustic responses were collected under various structural configurations over five days. Frequency-domain features were extracted using the Fast Fourier Transform (FFT), and classification was performed using five machine learning models: Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbors (KNN), XGBoost, and Multi-Layer Perceptron (MLP). Among these, MLP achieved the highest accuracy (71.4%), followed by SVM (65.7%), LightGBM (62.9%), KNN (60%), XGBoost (57.1%), and RF (51.4%). These results demonstrate the feasibility of diagnosing structural changes in robots using acoustic sensing alone, even with a compact hardware setup and limited training data. These findings suggest that acoustic sensing can provide a practical and efficient approach for robot fault diagnosis, offering potential applications in environments where conventional diagnostic tools are impractical. The study highlights the advantages of incorporating acoustic sensing into fault diagnosis systems and underscores its potential for developing accessible and effective diagnostic solutions for robotics. Full article

The dynamic process of eating—including chewing, biting, swallowing, food items, eating time and rate, mass, environment, and other metrics—may characterize behavioral aspects of eating. This article presents a systematic review of the use of sensor technology to measure and monitor eating behavior. The PRISMA 2020 guidelines were followed to review the full texts of 161 scientific manuscripts. The contributions of this review article are twofold: (i) A taxonomy of sensors for quantifying various aspects of eating behavior is established, classifying the types of sensors used (such as acoustic, motion, strain, distance, physiological, cameras, and others). (ii) The accuracy of measurement devices and methods is assessed. The review highlights the advantages and limitations of methods that measure and monitor different eating metrics using a combination of sensor modalities and machine learning algorithms. Furthermore, it emphasizes the importance of testing these methods outside of restricted laboratory conditions, and it highlights the necessity of further research to develop privacy-preserving approaches, such as filtering out non-food-related sounds or images, to ensure user confidentiality and comfort. The review concludes with a discussion of challenges and future trends in the use of sensors for monitoring eating behavior. Full article

Material removal processes such as milling, drilling, and turning often generate harmful vibrations that can negatively impact both machine performance and operator safety. Addressing these vibrations at their source or reducing them to safe levels is, therefore, a critical challenge. This study proposes a practical solution by introducing thin-fin-type vibration-absorbing devices fabricated using 3D printing technology. These devices are designed specifically to mitigate vibration propagation during milling operations. To evaluate their effectiveness, a multi-sensor system comprising sound level meters, a vibrometer, and a vision–acoustic camera was employed to measure sound levels. The results show that the use of fabricated devices can reduce noise levels significantly, from 93 dB (comparable to power tools or a lawn mower) to 74 dB (similar to normal conversation or a busy office). This substantial reduction demonstrates the potential of the proposed devices to enhance workplace safety and acoustic comfort on the shop floor. Full article

Autonomous driving technologies for environmental perception are mostly based on visual cues obtained from sensors like cameras, RADAR, or LiDAR. They capture the environment as if seen through “human eyes”. If this visual information is complemented with auditory information, thereby also providing “ears”, driverless cars can become more reliable and safer. In this paper, an Acoustic Event Detection model is presented that can detect various acoustic events in an automotive context along with their time of occurrence to create an audio scene description. The proposed detection methodology uses the pre-trained network Bidirectional Encoder representation from Audio Transformers (BEATs) and a single-layer neural network trained on the database of real audio recordings collected from different cars. The performance of the model is evaluated for different parameters and datasets. The segment-based results for a duration of 1 s show that the model performs well for 11 sound classes with a mean accuracy of 0.93 and F1-Score of 0.39 for a confidence threshold of 0.5. The threshold-independent metric mAP has a value of 0.77. The model also performs well for sound mixtures containing two overlapping events with mean accuracy, F1-Score, and mAP equal to 0.89, 0.42, and 0.658, respectively. Full article

This article examines the audiovisual representation of violence during the armed takeover of the Ecuadorian television channel TC Television on 9 January 2024, an unprecedented event in the country’s recent media history. Employing a film analysis methodology, the study deconstructs the live broadcast by segmenting it into visual sequences and analyzing elements such as narrative content, shot composition, camera movement, sound design, and editing techniques. The interpretive phase includes narratological, iconic, and psychoanalytic readings. From a psychoanalytic perspective, the study explores the emotional impact of the broadcast on viewers, focusing on responses such as fear, anxiety, identification, projection, and the activation of psychological defense mechanisms. It also reflects on the broader sociocultural consequences of such representations of violence in public media. The article concludes by emphasizing the need for public investment in inclusive and high-quality education as a structural response to youth vulnerability, school dropout, and the risk of recruitment by organized criminal groups in Ecuador. Full article

The characterization of dimensional stone deposits is essential for quarry assessment and design. However, uncertainties in mapping and designing pose significant challenges. To address this issue, an innovative approach is initiated to develop a virtual reality model by integrating unmanned aerial vehicle (UAV) photogrammetry for surface modeling and Electric Resistivity Tomography (ERT) for subsurface deposit imaging. This strategy offers a cost-effective, time-efficient, and safer alternative to traditional surveying methods for challenging mountainous terrain. UAV methodology involved data collection using a DJI Mavic 2 Pro (20 MP camera) with 4 K resolution images captured at 221 m altitude and 80 min flight duration. Images were taken with 75% frontal and 70% side overlaps. The Structure from Motion (SfM) processing chain generated high-resolution outputs, including point clouds, Digital Elevation Models (DEMs), Digital Surface Models (DSMs), and orthophotos. To ensure accuracy, five ground control points (GCPs) were established by a Real-Time Kinematic Global Navigation Satellite System (RTK GNSS). An ERT method known as vertical electric sounding (VES) revealed subsurface anomalies like solid rock mass, fractured zones and areas of iron leaching within marble deposits. Three Schlumberger (VES-1, 2, 3) and two parallel Wenner (VES-4, 5) arrays to a depth of 60 m were employed. The resistivity signature acquired by PASI RM1 was analyzed using 1D inversion technique software (ZondP1D). The integrated outputs of photogrammetry and subsurface imaging were used to design an optimized quarry with bench heights of 30 feet and widths of 50 feet, utilizing open-source 3D software (Blender, BIM, and InfraWorks). This integrated approach provides a comprehensive understanding of deposit surface and subsurface characteristics, facilitating optimized and sustainable quarry design and extraction. This research demonstrates the value of an innovative approach in synergistic integration of UAV photogrammetry and ERT, which are often used separately, for enhanced characterization, decision-making and promoting sustainable practices in dimensional stone deposits. Full article

Heterotrigona itama, a widely reared stingless bee species, produces highly valued honey. These bees naturally secure their colonies within logs, accessed via a single entrance tube, but remain vulnerable to intruders and predators. Guard bees play a critical role in colony defense, exhibiting the ability to discriminate between nestmates and non-nestmates and employing strategies such as pheromone release, buzzing, hissing, and vibrations to alert and recruit hive mates during intrusions. This study investigated the acoustic signals produced by H. itama guard bees during intrusions to determine their potential for intrusion detection. Using a Jetson Nano equipped with a microphone and camera, guard bee sounds were recorded and labeled. After preprocessing the sound data, Mel Frequency Cepstral Coefficients (MFCCs) were extracted as features, and various dimensionality reduction techniques were explored. Among them, Linear Discriminant Analysis (LDA) demonstrated the best performance in improving class separability. The reduced feature set was used to train both Support Vector Machine (SVM) and K-Nearest Neighbor (KNN) classifiers. KNN outperformed SVM, achieving a Precision of 0.9527, a Recall of 0.9586, and an F1 Score of 0.9556. Additionally, KNN attained an Overall Cross-Validation Accuracy of 95.54% (±0.67%), demonstrating its superior classification performance. These findings confirm that H. itama produces distinct alarm sounds during intrusions, which can be effectively classified using machine learning; thus, demonstrating the feasibility of sound-based intrusion detection as a cost-effective alternative to image-based approaches. Future research should explore real-world implementation under varying environmental conditions and extend the study to other stingless bee species. Full article

This paper is an attempt to investigate the rock indentation behaviors of a conical pick under different loading rates (1, 2, 3, and 4 mm/min), indenter types (sharp and blunt indenters), and types of rock (concrete, limestone, granite). Serial indentation tests by indenters were first performed by an automatic universal testing machine and monitored by an i-SPEED high-speed camera to record the peak pick force, indentation depth, rock fracture area, and rock failure process. Accordingly, the effect of loading rates, rock brittleness, and pick type on rock indentation behaviors was subsequently analyzed for a sound understanding of rock fragmentation mechanisms with indenters. It was found that higher loading rates necessitate a higher pick force and indentation depth to achieve rock fragmentation, resulting in a larger fractured area. Notably, a positive linear relationship exists between loading rates, rock-breaking forces, and fracture areas. A sharp indenter induces multiple cycles of repeated crushing and chipping phases, resulting in an arcuate-shaped fracture pattern with a smaller fractured area. Conversely, the rounded blunt indenter leads to a single stage of compression, with cracks propagating directly through the rock specimen, producing a larger fractured area. In addition, rock brittleness is another key factor to control rock failure efficiency, with tensile strength serving as a significant component. Full article

Magnetic resonance imaging (MRI) is a highly informative diagnostic method; however, its quality heavily depends on the patient’s immobility. Even minimal movements, such as breathing, can cause artifacts that complicate image interpretation, not to mention more significant movements, such as twitching or shivering. Given the high cost of the procedure, repeated scanning is undesirable. The aim of this study was to prepare patients for MRI procedures using specialized training software designed to minimize involuntary movements and improve diagnostic quality. The software tracked participants’ movements in an MRI simulator and reproduced characteristic scanning sounds. The Farnebäck optical flow algorithm detected even the slightest movements captured by the camera, allowing for the evaluation of movements during training sessions and improving patient readiness for actual scanning. A pilot study conducted on a group of 10 students aged 21–27 years demonstrated a significant reduction in the average number of movements during testing—from 27.7 in the first test to 8.3 in the second, corresponding to an average decrease of 19.4 movements. Additionally, two participants showed a noticeable reduction in anxiety levels after the first test, which likely contributed to the decrease in movements, emphasizing the importance of psychological preparation in enhancing training effectiveness. The study results suggest potential improvements in the quality of diagnostic images, which can increase their diagnostic value and enhance patient comfort during actual scanning, reducing the likelihood of repeated procedures. Full article

(1) Background: Human–wildlife conflict can lead to adverse consequences for both parties, particularly in areas with a high concentration of wild ungulates. Ungulates cause frequent, severe plant damage by stripping the bark or browsing on the youngest plants. In the latter case, they damage vegetative sprouts and leaves, which can cause a delay in growth or the plant’s death. Tuscany is notable for its significant population of wild boar, which cause substantial damage to vineyards and cereal crops, costing farmers millions annually. In Tuscany, given the highly cultivated landscape of olive trees, damage has also been recorded in these plants. Balancing human and wildlife needs is crucial for minimizing damage and ensuring coexistence. (2) Methods: This study tested innovative electronic playback devices using long-range radio technology (LoRa) to deter wild ungulates and prevent crop damage. These devices use sounds and lights to induce wild animals to be afraid and thus run away from the cultivated plot to be protected. The experiment was conducted on a farm in Chianti, Tuscany, involving four plots of land planted with olive trees: in two test areas, four playback devices and four camera traps were installed, and in the two control areas, only camera traps were installed. Playback devices aimed to deter wild ungulates and camera traps aimed to test their effectiveness. Data from the camera traps were analyzed statistically and behaviorally. (3) Results: Playback devices significantly reduced wild animal activity in the equipped areas. Statistical analysis revealed that the use of acoustic–luminous deterrent devices (PDs) significantly reduced wildlife visits to the olive groves. (4) Conclusion: The study’s findings, supported by heatmaps and frequency analyses, provide insights into wildlife activity patterns and guide the development of targeted, effective wildlife management strategies. Full article

The dynamic identification of the deformation of a noise barrier column is of great significance to the monitoring of its health. At the same time, the maximum stress of the column is an important indicator for the evaluation of its health status. Traditional contact displacement monitoring installs sensors on the structure, requires a lot of wiring and data acquisition equipment, and establishes a relatively independent and stable displacement reference system. Affected by the environment, wear, and material aging, the efficiency and reliability of data acquisition are reduced. A monitoring method based on digital image has the advantages of non-contact monitoring, high precision, and strong reliability. The existing DIC detection methods are limited by processor performance and image resolution, which are difficult to apply to engineering detection. In this paper, a structural displacement identification method based on convolutional neural networks (CNNs) and DIC technology is proposed. In this method, the data set is formed according to the column displacement cloud image obtained by DIC analysis, and the data set is enhanced by data normalization and region division. Through the analysis of the number of network layers and learning rate, the model design of the deep learning network is carried out. The high-speed camera image results of the test are introduced and identified by the static loading test of the equal-scale sound barrier. The results show that the structural displacement identification method based on CNN and DIC technology can accurately identify the displacement change in the structure, which greatly improves the efficiency of image displacement calculation using DIC technology. Full article