arXiv subjects

Francisco Roldan

Publications and source records attributed to Francisco Roldan.

3 recordsLinked to original sources

arXiv2022

The Lorentz factor in a reverse coordinate system

In the present study, we have derived the Lorentz factor using a coordinate system with antiparallel X-axes. Using a thought experiment, common in relativistic literature, we have used the case of a pulse of light moving along the X-axis. Next, we have argued that, consequently with the isotropy of space, the result must be the same if the trajectory of the light pulse is any angle $α>0^{\circ}$, thus obtaining an alternative transformation factor that generates the same results as the Lorenz factor at any angle of the trajectory of the light pulse, therefore confirming the relativistic isotropy of space in the context of the Minkowski spacetime. Nevertheless, a particularly important novelty is that the derived alternative transformation factor allows the reference frames involved to move at speeds greater than that of light.

physics.pop-ph↗

arXiv2021

An alternative transformation factor in the framework of the relativistic aberration of light

In the present study, we analyze in combination the principles of special relativity and the phenomenon of the aberration of light, deriving a system of equations that allows establishing the relationship between the angles commonly involved in this phenomenon. As a consequence, a transformation factor is obtained that generates two solutions, one of them with the same values as the Lorentz factor. This suggests that due to relativistic aberration, an apparent double image of celestial objects could be obtained. On the other hand, its functional form does not establish a limit to the speed of the reference frames with inertial movement.

physics.pop-ph↗

arXiv2019

Wav2Pix: Speech-conditioned Face Generation using Generative Adversarial Networks

Speech is a rich biometric signal that contains information about the identity, gender and emotional state of the speaker. In this work, we explore its potential to generate face images of a speaker by conditioning a Generative Adversarial Network (GAN) with raw speech input. We propose a deep neural network that is trained from scratch in an end-to-end fashion, generating a face directly from the raw speech waveform without any additional identity information (e.g reference image or one-hot encoding). Our model is trained in a self-supervised approach by exploiting the audio and visual signals naturally aligned in videos. With the purpose of training from video data, we present a novel dataset collected for this work, with high-quality videos of youtubers with notable expressiveness in both the speech and visual signals.

cs.MM↗