Néstor Becerra Yoma

To make the proper planning of bus public transportation systems, especially with the introduction of electric buses to the fleets, it is essential to characterize the routes, patterns of traffic, speed, constraints, and presence of high slopes. Currently, GPS (Global Position System) is available worldwide in the fleet. However, they often produce datasets of poor quality, with low data rates, loss of information, noisy samples, and eventual paths not belonging to regular bus routes. Therefore, extracting useful information from these poor data is a challenging task. The current paper proposes a novel method based on an unsupervised competitive density clustering algorithm to obtain hot spot clusters of any density. The clusters are a result of their competition for the GPS samples. Each cluster attracts GPS samples until a maximum radius from its centroid and thereafter moves toward the most density areas. The winning clusters are sorted using a novel distance metric with the support of a visual interface, forming a sequence of points that outline the bus trajectory. Finally, indicators are correlated to the clusters making a trajectory characterization and allowing extensive assessments. According to the actual case studies, the method performs well with noisy GPS samples and the loss of information. The proposed method presents quite a fixed parameter, allowing fair performance for most GPS datasets without needing custom adjustments. It also proposes a framework for preparing the input GPS dataset, clustering, sorting the clusters to outline the trajectory, and making the trajectory characterization.

Human–robot collaboration will play an important role in the fourth industrial revolution in applications related to hostile environments, mining, industry, forestry, education, natural disaster and defense. Effective collaboration requires robots to understand human intentions and tasks, which involves advanced user profiling. Voice-based communication, rich in complex information, is key to this. Beamforming, a technology that enhances speech signals, can help robots extract semantic, emotional, or health-related information from speech. This paper describes the implementation of a system that provides substantially improved signal-to-noise ratio (SNR) and speech recognition accuracy to a moving robotic platform for use in human–robot interaction (HRI) applications in static and dynamic contexts. This study focuses on training deep learning-based beamformers using acoustic model-based multi-style training with measured room impulse responses (RIRs). The results show that this approach outperforms training with simulated RIRs or matched measured RIRs, especially in dynamic conditions involving robot motion. The findings suggest that training with a broad range of measured RIRs is sufficient for effective HRI in various environments, making additional data recording or augmentation unnecessary. This research demonstrates that deep learning-based beamforming can significantly improve HRI performance, particularly in challenging acoustic environments, surpassing traditional beamforming methods.

This paper presents a unified model for combining beamforming and blind source separation (BSS). The validity of the model's assumptions is confirmed by recovering target speech information in noise accurately using Oracle information. Using real static human-robot interaction (HRI) data, the proposed combination of BSS with the minimum-variance distortionless response beamformer provides a greater signal-to-noise ratio (SNR) than previous parallel and cascade systems that combine BSS and beamforming. In the difficult-to-model HRI dynamic environment, the system provides a SNR gain that was 2.8 dB greater than the results obtained with the cascade combination, where the parallel combination is infeasible.

A respiratory distress estimation technique for telephony previously proposed by the authors is adapted and evaluated in real static and dynamic HRI scenarios. The system is evaluated with a telephone dataset re-recorded using the robotic platform designed and implemented for this study. In addition, the original telephone training data are modified using an environmental model that incorporates natural robot-generated and external noise sources and reverberant effects using room impulse responses (RIRs). The results indicate that the average accuracy and AUC are just 0.4% less than those obtained with matched training/testing conditions with simulated data. Quite surprisingly, there is not much difference in accuracy and AUC between static and dynamic HRI conditions. Moreover, the beamforming methods delay-and-sum and MVDR lead to average improvement in accuracy and AUC equal to 8% and 2%, respectively, when applied to training and testing data. Regarding the complementarity of time-dependent and time-independent features, the combination of both types of classifiers provides the best joint accuracy and AUC score.

In this paper, a system to assess dyspnea with the mMRC scale, on the phone, via deep learning, is proposed. The method is based on modeling the spontaneous behavior of subjects while pronouncing controlled phonetization. These vocalizations were designed, or chosen, to deal with the stationary noise suppression of cellular handsets, to provoke different rates of exhaled air, and to stimulate different levels of fluency. Time-independent and time-dependent engineered features were proposed and selected, and a k-fold scheme with double validation was adopted to select the models with the greatest potential for generalization. Moreover, score fusion methods were also investigated to optimize the complementarity of the controlled phonetizations and features that were engineered and selected. The results reported here were obtained from 104 participants, where 34 corresponded to healthy individuals and 70 were patients with respiratory conditions. The subjects’ vocalizations were recorded with a telephone call (i.e., with an IVR server). The system provided an accuracy of 59% (i.e., estimating the correct mMRC), a root mean square error equal to 0.98, false positive rate of 6%, false negative rate of 11%, and an area under the ROC curve equal to 0.97. Finally, a prototype was developed and implemented, with an ASR-based automatic segmentation scheme, to estimate dyspnea on line.

In this letter, a method based on long short-term memory (LSTM) is presented to address the problem of earthquake magnitude estimation for earthquake early warning (EEW) and tsunami early warning (TW) purposes using a seismic station. An end-to-end-based scheme is adopted, and particular attention is paid to compute the magnitude of seismic events larger than M6 and reduce the effective time for TW. More so, these earthquakes are the ones that cause more fear or uncertainty in the population with the provision of the most significant destructive potential. However, the occurrence of large earthquakes is low, but to counteract the drawback of limited training data, engineered features were also proposed. The earthquake magnitude relative error estimation reported here in experiments with Chilean seismic data was 4.01% and 8.04% with earthquakes M4.0 or larger (up to M8.1) and M4.0 or smaller, respectively, by employing seismic traces in the nearest station to the corresponding seismic event. The average earthquake-nearest station distance was 196 km, and in 26% of the data, this distance was greater than 200 km. These results are competitive with those published elsewhere and suggest the possibility to reduce the time required for EEW and especially TW.