VTVan-Thuan Tran

Van-Thuan Tran

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Van-Thuan Tran has a total of 20 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    On the Object Recognition via Free-Fall Impact Vibration Signals

    by Chin-Chieh Chang, Van-Thuan Tran, Jung-Tsung Kuo, Wei-Ho Tsai

    Object recognition through physical interaction provides an alternative sensing approach for robotic systems and intelligent devices when visual information is unavailable or insufficient. This paper investigates object recognition using free-fall impact vibration signals. We collect vibration data using an inertial sensor module measuring acceleration and angular velocity during object impacts on a spring-mounted wooden platform, and evaluate our approach on five datasets with increasing difficulty: 10 objects with distinct weights (Set 1), 15 objects with similar weights but different shapes (Set 2), 20 highly similar screwdriver bits (Set 3), a unified 45-object dataset, and an extended 60-object dataset. We propose a Temporal Convolutional Network (TCN) architecture and examine various approaches to process acceleration and angular velocity signals along the vertical axis. Session-based 10-fold cross-validation, where each fold corresponds to a complete data collection session, is used to ensure evaluation of real-world generalization capability. Experimental results demonstrate that TCN significantly outperforms baseline architectures (1D-CNN, CNN-LSTM, SVM) across all datasets and experiments. On the challenging unified dataset, single-sensor TCN achieves 94.27% accuracy with angular velocity data, while dual-sensor fusion achieves 94.51%. TCN maintains robust performance from Set 1 to Set 3, demonstrating practical applicability even with highly similar objects. Remarkably, TCN achieves this performance with the parameter efficiency (0.6M parameters for single-branch, 1.2M for dual-branch), outperforming 1D-CNN dual-branch (8.3M parameters, 92.80% accuracy) with much fewer parameters. Analysis also reveals that angular velocity consistently outperforms acceleration by 1.71% to 15.69% depending on the model and validates the effectiveness of TCN for impact vibration classification with moderate training data. The method demonstrates robustness across signal lengths (1-2s optimal) and maintains low cross-session variance (1-2%), indicating stable performance across environmental conditions.

  • Open Access

    Cross-Device Nonverbal-Based Person Identification Using Mobile-Sensed Breath Sounds

    by Jung-Tsung Kuo, Van-Thuan Tran, Lydia Hsiao-Mei Lin, Yuh-Shyan Hwang, Wei-Ho Tsai

    Person identification using breath sounds has emerged as a promising biometric modality. While previous research demonstrated the feasibility of using stethoscopes or single mobile devices, the critical challenge of cross-device generalization remains unexplored. This work presents the investigation of cross-device person identification using mobile-sensed bronchial breath sounds. We collected a multi-device dataset comprising recordings from 20 subjects using three different mobile devices, with three sessions per subject per device. We propose a framework combining the Audio Spectrogram Transformer (AST) with supervised contrastive learning and dual data augmentation (traditional techniques and diffusion-based augmentation). Evaluation using device-based and session-based cross-validation reveals that cross-device scenarios are significantly more challenging than same-device scenarios (a 5-7% gap in baseline). Our proposed method achieves 85.99-88.99% cross-device accuracy and 90.85-92.58% same-device accuracy, significantly narrowing the performance gap. The AST demonstrates clear superiority over CNN baselines (VGGish, ResNet50) with 16-25% advantages. Ablation studies confirm that supervised contrastive learning and data augmentation each contribute 2-4% improvement, with synergistic benefits when combined. Robustness evaluation shows the proposed method maintains 82-87% accuracy even at 0dB SNR, while the baseline degrades dramatically. Few-shot learning experiments demonstrate remarkable data efficiency, achieving comparable performance with only 5 segments per session. These findings establish the feasibility of practical cross-device breath-sound biometrics with well-acceptable noise robustness and minimal enrollment requirements.

  • Open Access

    On the Cross-Container Recognition of Liquid Types From Pouring Sounds

    by Van-Thuan Tran, Chin-Chieh Chang, Jung-Tsung Kuo, Yu-Xin Tien, Wei-Ho Tsai

    Identifying liquid types through acoustic signatures offers promising opportunities for food service quality control, dietary monitoring, and assistive technologies. Unlike prior acoustic liquid sensing approaches that rely on active signal transmission or specialized hardware, this work addresses a promising alternative: classifying liquids from passive pouring sounds using only a consumer microphone across diverse container types. The main contributions are: (1) the formal framing of passive Acoustic-based liquid recognition (ALR) as a domain generalization problem with systematic leave-one-container-out (LOCO) evaluation; (2) a multi-task learning framework combining the Audio Spectrogram Transformer (AST) with supervised contrastive and adversarial training for container-invariant recognition; and (3) comprehensive evaluation across seven containers spanning three material categories. The evaluation begins with train-on-one-container-only (TOCO) experiments that expose severe generalization challenges: models trained on single containers achieve only 59.27% mean accuracy on unseen containers, with AST outperforming CNN baselines (VGGish, ResNet-50). Under LOCO evaluation on seven containers (ceramic, glass, and paper cups, 170-390ml), our proposed AST with multi-task learning achieves 95.02% mean accuracy, significantly exceeding baseline AST training (85.80%), with ablation studies revealing complementary contributions of supervised contrastive learning and adversarial training. Performance analysis across varying segment durations (2-10 seconds) reveals recognition capability even with severely constrained temporal information: 5-second segments attain 92.28% accuracy, 3-second segments reach 90.14%, and 2-second segments achieve 88.45%, the latter still exceeding baseline performance at full duration. These results validate that container-invariant ALR is attainable through domain-invariant representation learning, enabling future development for practical non-contact liquid identification systems deployable across arbitrary container types.

  • Open Access

    Pouring Sound-Based Cross-Container Water Temperature Recognition Using Single-Container Training

    by Yi-Li Chen, Yu-Wei Kuo, Van-Thuan Tran, Wei-Ho Tsai

    Acoustic-based temperature recognition (ATR) for flowing water offers a non-contact sensing alternative for various applications. However, practical deployment faces a critical challenge: training data is often available from only a single reference container, yet the system must operate across diverse containers in real-world use. Container-specific acoustic characteristics, such as resonance modes, material properties, and geometric effects, can dominate recorded pouring sounds and prevent direct transfer of models trained on one container to others. We tackle this single-container training scenario using the Audio Spectrogram Transformer (AST) as the base model, combining data augmentation (DA) with cross-entropy (CE) objective and supervised contrastive learning (SupCon) to learn container-invariant temperature representations. To contextualize our approach, we first evaluate several unsupervised source separation (USS) methods for extracting temperature-related sources as potential solutions; however, all separation approaches underperform the original, unseparated data, indicating that explicit source decomposition is ineffective for this task. Our proposed DA+CE+SupCon framework does not require explicit source decomposition or ground truth separated signals; instead, it leverages data augmentation and contrastive loss to enforce similarity between same-temperature samples regardless of container identity. Evaluating this approach on a dataset of four containers (two ceramic, two glass) across three temperature ranges (cold 20-25°C, warm 50-55°C, hot 80-95°C), we demonstrate significant improvements in cross-container transferability within the evaluated material set. When training on a single container with full data and testing on others, our approach achieves 4-10% higher cross-container accuracy than models trained with standard cross-entropy loss. In extreme data-scarcity scenarios (5-shot learning), the proposed approach further yields 13-23% gains, demonstrating exceptional robustness even with severely limited training data. These results enable practical ATR deployment with minimal training data collection requirements across the evaluated container types; generalization to other materials remains unexplored and is discussed as future work.

  • Open Access

    Automatic Identification of Textile Material From Cutting Sounds

    by Zong-Fu Yang, Van-Thuan Tran, Hsu-Hui Chan, Wei-Ho Tsai

  • Listening to Temperature: A Novel Temperature Measurement for Flowing Water Using Audio Recognition Techniques

    by Van-Thuan Tran, Po-Hsuan Shang, Yi-Li Chen, Wei-Ho Tsai

    This study proposes an innovative temperature measurement technique for flowing water, which uses the sound of pouring water to determine its temperature, thereby providing a non-contact temperature sensing technology. It is found that when water is poured from one container into another, different water temperatures cause subtle differences in the pouring sound. Therefore, we can use a microphone to record the sound generated during the pouring process and then determine the temperature of the water as it falls based on the recorded sound. Compared to most contact-based temperature sensors currently used to measure liquid temperature, this technique has the advantage of not causing contact contamination or affecting the water temperature. It can also be designed as an App on a handheld mobile device, where the built-in microphone records the sound to recognize the water temperature, thus eliminating the need for additional hardware costs. To achieve this, we constructed a dataset of pouring sounds across three temperature classes (cold, warm, and hot) and extracted acoustic features such as MFCCs (Mel-frequency cepstral coefficients) and Mel-spectrograms. A deep learning model, ATRNet (Audio-based Temperature Recognition Network), was proposed and benchmarked against CRNN (Convolutional Recurrent Neural Network), VGGish architecture, and SVM (Support Vector Machines) using different input durations and noise conditions. Experimental results show that ATRNet with combined MFCCs and Mel-spectrogram features achieves state-of-the-art performance, up to 100% accuracy on clean post-full segments and strong robustness under noise. Even with short audio segments (3s and 5s), the system maintains competitive accuracy, demonstrating its potential for real-time and embedded applications. These results confirm the feasibility of using impact sound during water pouring for practical, non-contact temperature recognition.


  • Open Access

    Non-Contact Hand Tremor Detection Via Water Pouring Sound Recognition

    by Van-Thuan Tran, Thi-Tho Nguyen, Yi-Li Chen, Wei-Ho Tsai

  • Open Access

    On the Acoustic-Based Recognition of Multiple Objects Using Overlapped Impact Sounds

    by Van-Thuan Tran, Wei-Ho Tsai

    In a recent study, we have uncovered the potential of utilizing impact sounds generated when objects fall freely and strike a plane for acoustic-based object recognition (AOR). Building upon this discovery, we address the practical scenario where multiple objects fall simultaneously, leading to overlapped impact sounds. This investigation marks the inception of our work on recognizing multiple objects using overlapped impact sounds, termed acoustic-based multiple object recognition (AMOR). To tackle the challenges posed by overlapped sounds in AOR, we propose a novel two-branch network named OANet, designed with overlap awareness. This network integrates transformer encoding blocks, with the first branch pre-trained using the supervised contrastive learning approach to predict the number of objects. The outputs of the first and second branches are combined and fed into a classifier for the AMOR task. We conducted experiments and evaluations on a self-collected dataset of 10 distinct LEGO objects, including instances where a maximum of three objects fell and hit the designated plane almost concurrently. This dataset contains both non-overlapped and overlapped samples across 175 classes or object combinations (i.e., 10 individual objects, 45 combinations of two objects, and 120 combinations of three objects). Experimental results underscore the viability and promising outcomes of AMOR, showcasing the potential for accurate recognition in complex scenarios. Notably, OANet exhibits superior performance and robustness compared to baseline networks without overlap awareness. Furthermore, upon investigating AMOR in more complex scenarios with mixed overlapped datasets containing samples of a maximum of 5, 7, and 10 objects, we observed the remarkable efficiency of the proposed OANet. It consistently achieved accuracies surpassing 91%, which were 14-24% higher than the results obtained by the baseline models. Additionally, other important findings include the effectiveness of feature combination, the supervised contrastive learning approach for feature learning in the overlap detection network, and the use of mixed data during the training phase.

  • Open Access

    Acoustic-based Recognition of Multiple Objects with Overlapped Impact Sounds

    by Ting-Hao You, Van-Thuan Tran, Chun-Shuo Huang, Chia-Yang Wu, Thi-Tho Nguyen, Wei-Ho Tsai

    Impact sounds resulting from object interactions with the environment or other objects are valuable for acoustic-based object recognition (AOR). However, existing research has primarily focused on single-object recognition with non-overlapped impact sounds. This work investigates the practical scenario in AOR where multiple objects can generate impact sounds simultaneously, leading to overlapped sounds. This investigation marks the inception of our work on recognizing multiple objects using overlapped impact sounds, termed acoustic-based multiple object recognition (AMOR). To facilitate experimentation in this pilot investigation, we collected datasets comprising non-overlapped impact sounds from 10 LEGO objects and created four datasets with overlapped instances of up to 3, 5, 7, and 10 objects using the data-mixing approach. To handle AOR covering both non-overlapped and overlapped sounds, we propose OANet, a two-branch network designed with overlap awareness. OANet integrates transformer encoding blocks, with one branch pre-trained to predict object numbers. Experimental results demonstrate the viability and promising outcomes of AMOR, with OANet outperforming baseline networks. In complex scenarios with mixed overlapped datasets containing samples of up to 5, 7, and 10 objects, OANet consistently achieves accuracies exceeding 91%, surpassing baseline models by at least 14%.