In this paper, we present a new single sound source DOA estimation and tracking system based on the well-known SRP-PHAT algorithm and a three-dimensional Convolutional Neural Network. It uses SRP-PHAT power maps as input features of a fully convolutional causal architecture that uses 3D convolutional layers to accurately perform the tracking of a sound source even in highly reverberant scenarios where most of the state of the art techniques fail. Unlike previous methods, since we do not use bidirectional recurrent layers and all our convolutional layers are causal in the time dimension, our system is feasible for real-time applications and it provides a new DOA estimation for each new SRP-PHAT map. To train the model, we introduce a new procedure to simulate random trajectories as they are needed during the training, equivalent to an infinite-size dataset with high flexibility to modify its acoustical conditions such as the reverberation time. We use both acoustical simulations on a large range of reverberation times and the actual recordings of the LOCATA dataset to prove the robustness of our system and its good performance even using low-resolution SRP-PHAT maps.

In this paper, we present a new model for Direction of Arrival (DOA) estimation of sound sources based on an Icosahedral Convolutional Neural Network (CNN) applied over SRP-PHAT power maps computed from the signals received by a microphone array. This icosahedral CNN is equivariant to the 60 rotational symmetries of the icosahedron, which represent a good approximation of the continuous space of spherical rotations, and can be implemented using standard 2D convolutional layers, having a lower computational cost than most of the spherical CNNs. In addition, instead of using fully connected layers after the icosahedral convolutions, we propose a new soft-argmax function that can be seen as a differentiable version of the argmax function and allows us to solve the DOA estimation as a regression problem interpreting the output of the convolutional layers as a probability distribution. We prove that using models that fit the equivariances of the problem allows us to outperform other state-of-the-art models with a lower computational cost and more robustness, obtaining root mean square localization errors lower than $10^{\circ }$ even in scenarios with a reverberation time $\mathbf {T_{60}}$ of $1.5 \,\mathrm{s}$ .

Automatic music transcription (AMT) is a critical problem in the field of music information retrieval (MIR). When AMT is faced with deep neural networks, the variety of timbres of different instruments can be an issue that has not been studied in depth yet. The goal of this work is to address AMT transcription by analyzing how timbre affect monophonic transcription in a first approach based on the CREPE neural network and then to improve the results by performing polyphonic music transcription with different timbres with a second approach based on the Deep Salience model that performs polyphonic transcription based on the Constant-Q Transform. The results of the first method show that the timbre and envelope of the onsets have a high impact on the AMT results and the second method shows that the developed model is less dependent on the strength of the onsets than other state-of-the-art models that deal with AMT on piano sounds such as Google Magenta Onset and Frames (OaF). Our polyphonic transcription model for non-piano instruments outperforms the state-of-the-art model, such as for bass instruments, which has an F-score of 0.9516 versus 0.7102. In our latest experiment we also show how adding an onset detector to our model can outperform the results given in this work.

The application of MEMS capacitive accelerometers is limited by its thermal dependence, and each accelerometer must be individually calibrated to improve its performance. In this work, a light calibration method based on theoretical studies is proposed to obtain two characteristic parameters of the sensor’s operation: the temperature drift of bias and the temperature drift of scale factor. This method requires less data to obtain the characteristic parameters, allowing a faster calibration. Furthermore, using an equation with fewer parameters reduces the computational cost of compensation. After studying six accelerometers, model LIS3DSH, their characteristic parameters are obtained in a temperature range between 15 °C and 55 °C. It is observed that the Temperature Drift of Bias ($TDB$) is the parameter with the greatest influence on thermal drift, reaching 1.3 mg/°C. The Temperature Drift of Scale Factor ($TDSF$) is always negative and ranges between 0 and −400 ppm/°C. With these parameters, the thermal drifts are compensated in tests with 20 °C of thermal variation. An average improvement of 47% was observed. In the axes where the thermal drift was greater than 1 mg/°C, the improvement was greater than 80%. Other sensor behaviors have also been analyzed, such as temporal drift (up to 1 mg/h for three hours) and self-heating (2–3 °C in the first hours with the corresponding drift). Thermal compensation has been found to reduce the effect of the latter in the first hours after power-up of the sensor by 43%.