Researcher

Expires on:With Published chosen, won't be shown after this time Ircam is recruiting for its Sound Analysis and Synthesis Department a Researcher, on a full-time fixed-term contract of 18 months, with the possibility of extension to 36 months, starting in September 2026. INTRODUCTION TO IRCAM IRCAM is a non-profit organization, associated with the Centre National d'Art et de Culture Georges Pompidou, whose missions include research, creation and teaching activities around 20th century music and its relationship with science and technology. Within the joint research unit, UMR 9912 STMS (Sciences et Technologies de la Musique et du Son), affiliated with IRCAM, Sorbonne University, CNRS and the French Ministry of Culture and Communication, specialized teams carry out research and development work in acoustics, sound signal processing, cognitive sciences, interaction technologies, computer music and musicology. IRCAM is located in the centre of Paris, near the Centre Georges Pompidou at 1, Place Stravinsky 75004 Paris, France.The work will be carried out at IRCAM within the Sound Analysis and Synthesis team (A/S), which is specialized in sound synthesis and transformation, under the supervision of Nicolas Obin and Axel Roebel. PROJECT DESCRIPTION AI-MADE, a research project conducted in the STMS laboratory, is part of the PostGenAI@Paris Cluster, headed by Sorbonne Université and operated by the French National Research Agency (ANR) under the France 2030 programme, with the ambition of creating a pole of excellence in next-generation AI. Within this context, AI-MADE aims to develop innovative generative audio models extending the capacities of human creators to transform artistic ideas into high-quality audio content, by significantly expanding the limits of creative feasibility and by reducing the time and effort required for professional output. The core scientific challenge to be investigated is the intuitive and precise control of the perceptual properties of the generated sounds. The research conducted has direct and various applications in music production, live performance, cinema, or video games. Addressing the large computational costs and negative environmental impact of advanced generative models, AI-MADE will favour models with properly adapted inductive bias, thereby achieving improved computational efficiency, and reduced data requirements. Simultaneously, AI-MADE will aim for an appropriate organization of latent representations fostering intuitive manipulation by human creators. For selected problems, AI-MADE will explore the development of low latency models that may be used in real-time. Finally, AI-MADE will implement functional prototypes that will be experimented directly in the context of audio productions and artistic residencies. JOB DESCRIPTION The candidate is expected to actively participate in all activities related to the AI-MADE project (research, teaching, applications). Research: The candidate will actively contribute to the animation and the supervision of the research activities (internships, PhD theses) in coordination with the A/S members involved in the project. He/She will contribute to original research in audio generative models with a particular focus on the intuitive controllability of the generation, the definition of controls which responds to actual needs of artistic applications and professional productions, and the computational and data efficiency of the generative models. In complement to the research, he/she will also assist the integration of his/her into the development of functional prototypes. Finally, he/she will be protagonist in the animation of the two artistic residencies and participate to the experimentation of the produced generative audio models with artists and professionals. Teaching: The candidate will dispense 64 hours of teaching per year at Sorbonne University. He/She is expected to develop teaching activities aimed at promoting and disseminating the research conducted within AI-MADE to a diverse audience, ranging from students to expert users, through various formats, including university courses, workshops, spring and/or summer schools, hackathons and/or creathons. The lectures will mainly take place in the Faculty of Science and Engineering at Sorbonne University, the Sorbonne Cluster for Artificial Intelligence (SCAI), and possibly other partners of the PostGenAI@Paris cluster. The topics aim to cover all scientific domains that are relevant for generative modelling: computer science (learning algorithms), signal processing (speech and music), computational linguistics (language models), and will develop original educational content that enrich existing computer science programs addressing practical, creative, legal, and ethical (privacy, intellectual property, ...) aspects. The audience are possibly students (computer sciences, humanities, social sciences, law, the arts), but also artists, and professionals from creative industries. EXPERIENCE AND SKILLS REQUIRED The candidate must have a professional experience of associate professor, lecturer, and/or post‑doc. A PhD Thesis and a preliminary experience in audio generative models and creative applications is wanted. Proficiency in audio generative models, in particular of deep neural networks, demonstrated by significant scientific contribution in the domain; Strong expertise of digital audio signal processing, in particular to speech and music signal processing; Excellent knowledge of the Python programming language, the TensorFlow and/or PyTorch frameworks, and distributed computing on #J-18808-Ljbffr