Machine Listening, or AI for Sound, is defined as the general field of Artificial Intelligence applied to audio analysis, understanding and synthesis by a machine. The access to ever increasing super-computing facilities, combined with the availability of huge data repositories (although largely unannotated), has led to the emergence of a significant trend with pure data-driven machine learning approaches. The field has rapidly moved towards end-to-end neural approaches which aim to directly solve the machine learning problem for raw acoustic signals but often only loosely taking into account the nature and structure of the processed data. The main consequences are that the models are 1) overly complex, require massive amounts of data to be trained and extreme computing power to be efficient (in terms of task performance), and 2) remain largely unexplainable and non-interpretable. To overcome these major shortcomings, we believe that our prior knowledge about the nature of the processed data, their generation process and their perception by humans should be explicitly exploited in neural-based machine learning frameworks. The aim of HI-Audio is to build such hybrid deep approaches combining parameter-efficient and interpretable signal models, musicological and physics-based models, with highly tailored, deep neural architectures. The research directions pursued in HI-Audio will exploit novel deterministic and statistical audio and sound environment models with dedicated neural auto-encoders and generative networks and target specific applications including speech and audio scene analysis, music information retrieval and sound transformation and synthesis.

Mechanomics refers to the measurement of mechanical properties at the microscopic scale in biological tissues. Strong hopes are currently placed on mechanomics to evaluate quantitatively how a treatment or a gene expression affects the stiffness or strength of a tissue, with major impacts expected in drug discovery, diagnostics and genomics screening. However, there is a pressing need for new instrumentation technologies in mechanomics. In the BIOLOCHANICS ERC CoG project, we developed and validated a novel multimodal technology addressing these challenges. Our technology can perform the following actions: 1. apply controlled loads on tissue samples, 2. measure the induced bulk deformations at the micron level and 3. map the distribution of local stiffness of these tissues. Our technology shows very competitive potential for mechanomics in general. However, as an innovative technology, it remains now at the stage of a technological concept with a first laboratory application (TRL2-3) achieved within the ERC CoG BIOLOCHANICS project. Our global objective in MECHANOMICS-POC is to reach TRL 6 in order to take it further towards a commercial innovation: make a first integrated prototype, test it in an intended environment, and refine the comparison with existing technologies. Our specific objectives are (O1) to build an integrated user-friendly prototype; (O2) to determine further IPR strategy; (O3) to establish the commercialization strategy including market research, industrial partnerships and lead management; (O4) to validate the prototype in real-world conditions and present it to industrial stakeholders in the identified sectors. Beside the development and assessment of the prototype demonstrator, we will focus on the future exploitation and commercialization roadmap of this technology, which is aimed at becoming a standard laboratory equipment or service for fundamental research in life sciences and for drug discovery applications.

Opinion mining is a progressing domain. A lot of efforts have been recently dedicated to the development of methods able to analyze opinion data available on the social Web. At the same time, companies that are developing companion robots and virtual vocal assistants (Siri, Google Now, Cortana, etc.) show a growing interest for the integration of the social component in the interaction. MAOI is a fundamental research project in natural language processing and speech processing which contributes to Challenge 7 (Society of information and communication). It deals with opinion analysis in human-agent interaction and is thus integrated in Axis 4 (Interactions, robotics). More precisely, the MOAI project tackles multimodal opinion analysis methods in human-agent multimodal interactions in order to extract information concerning user’s preferences. Such information is dedicated to enrich user profiles for companion robots and virtual assistants. This challenging issue has been so far rarely and partially handled by the state of the art. The proposed approach relies on Conditional Random Fields (CRF) that have been chosen for their flexibility in order to take advantage of both the generalization capability of machine learning methods and the fine-grain modeling of semantic rules. As recordings of face-to face human-agent interactions are not yet massively available, such flexible methods constitute an alternative to deep learning methods. In this promising context, the MAOI project targets two major breakthroughs: i) feature learning driven by a priori knowledge and psycho-linguistic models in order to learn users' preferences; ii) the integration of various levels of analysis (lexical, syntactic, prosodic, dialogic) through latent variables inside hidden CRF, allowing for grounding the opinion detection in the context of human-agent interaction.

Although affordable solutions exist to store and recover heat from low temperature sources, the high temperature ones (> 550 °C) such as the concentrated solar and waste heat from high temperature energy intensive industries remain challenging since efficient and affordable storage materials are scarce. The waste heat from these industries is huge and corresponds to 16% (122 Terawatt hours) of the total heat consumption/year in Europe. STOREHEAT targets the investigation of an outstanding and novel family of High Storage Capacity materials, namely Calcium Carbide-based Composites (3C), for High Temperature Heat Storage. 3C is produced at much lower temperature (1000 – 1200 °C) than the current solutions (1800-2500 °C) based on silicon carbide (SiC) ceramics from fossil source (coke) mainly. 3C is synthesized by carbonization of calcium rich biochar and have not yet been mentioned in the literature neither for high temperature storage nor the mechanism of their formation explored. Preliminary synthesis attempts showed a high temperature storage capacity for 3C of 20% higher than that of SiC while consuming much less energy and using sustainable resources for its production. The scientific breakthrough lies on the counter-intuitive and pioneering approach proposed to combine and stabilize the hierarchical carbon and metal species both from biochar to take advantage of their respective high thermal conductivity and heat capacity at high temperature. This seemingly winning combination, key for effective heat storage has never been done before. To achieve this objective, I propose an ambitious research approach combining in-situ and dynamic experimental methods and modelling to unlock the mechanisms governing the chemical phases assemblage and stabilization of 3C. The storage performance will be evaluated and optimized. The findings will push a way beyond the frontier of knowledge and broaden research opportunities in scientific communities interested in energy storage.