VALENCIA, Oct. 31 (EUROPA PRESS) –
Researchers from ValgrAI – an alliance between companies, the Generalitat and the five Valencian public universities – have presented their innovation projects in artificial intelligence (AI) with which they want to raise the Valencian Community as a reference: from tools to detect hoaxes, to clone human voices with which to achieve automatic utterances to the modeling of socioeconomic systems and other scenarios.
The Second Meeting of the Joint Research Unit of ValgrAI held in La Nau has had the participation of the general director of ValgrAI, Vicente Botti, and the managing director, Ana Cidad, who has presented the Artificial Intelligence Innovation Alliance (AIIA) that aims generate an innovation community with researchers, students and professionals from universities, research centers, companies and entities, as explained by ValgrAI in a statement.
The first presentation of the day was given by Emilio Soria-Olivas, representative of the Electronic Industrial Engineering Department of the Higher Technical School of Engineering of the University of Valencia, who addressed ‘Industrial AI Projects in IDAL’.
Specifically, he has exposed the predictive behavior models applied to Aranco and Galp, Voclity, the use of ‘text to speech’ models to clone human voices and be used in the automatic generation of speeches, distortion in high-power speakers for the company DAS, insect detection for Dismutel and identification of fake IDs for FOXid. A wide range of examples that show the broad possibilities of AI for companies.
The researcher Paolo Rosso from the Universitat Politècnica de València has shown “areas of natural language processing in PRHLT”. Specifically, he focused on the application of AI in the detection of fake news, conspiracy theories and hate, sexist or misogynistic messages.
In this sense, he has stated that these issues are of great concern to Spaniards. In fact, before the coronavirus, 88% thought that these issues were a problem, and currently, 86% of citizens consider them to be issues that must be addressed.
Rosso has highlighted the importance of emotions when it comes to issuing this type of messages that are harmful to society and has referred to a recent EMT study where it is stated that ‘fake news’ spreads more quickly than true information because Emotions are appealed to, not only through the lexicon but also through the image. Thus, mechanisms are being established to detect and analyze this type of behavior, such as the Hurtful HUmour (HUHU) Detection of humor spreading prejudice on Twitter.
On the other hand, the professor of Constitutional Law at the University of Valencia, Lorenzo Cotin, gave the presentation ‘Regulation and rights of AI’, in which he explained that he not only directs several research projects from the Faculty of Law related to with AI, but currently in this faculty there are more than 30 professors dedicated to digital law and AI and very cutting-edge research groups, which make the Valencian Community a leading autonomy in these topics.
In addition, he has pointed out that, although work is being done on an AI regulation that companies must already take into account, although it takes two years to be applied, there is regulation on AI because there are already laws such as the Data Protection Law and Law. to Privacy that are mandatory.
The professor of the Department of Electronic Engineering and representative of the Intelligent Data Analysys Laboratory (IDAL), José Martín Guerrero, from the University of Valencia, has spoken about Quantum Artificial Intelligence and has delved into the projects Quantum memristors, PINNs for counterdabatic driving, Quantum information retrieval and Quantum clustering.
Raul Montoliu, representative of the UJI, has shown the project of ‘creating agents for multi-action games’. At the university they have devised ASMACAG (A simple multi-action card game) to establish reinforcement learning algorithms, generate understandable and well-structured code and check if it works.
Carlos Marín, one of the post-doctoral researchers funded by the ValgrAI Foundation and member of the Gamers research group at the UJI, has addressed the AI ??tools applied to the development of video games, stating that currently automation through AI The processes between illustrators and programmers should have more weight to optimize processes.
Finally, from the Computer Technology Group of the Polytechnic University, Joaquín Taverner, has presented an initiative that aims to lay the foundations for Responsible AI through the modeling of complex socioeconomic and industrial systems, solutions based on automation using AI and the simulation of possible scenarios. All this with the aim of generating a virtual society in which humans and agents can coexist.