Gustav Eje Henter

Gustav Eje Henter is an Assistant Professor at KTH Royal Institute of Technology, holding roles as the Head of Research at Motorica AB and a Core Team Member of the Wallenberg Research Arena (WARA) for Media and Language. He is the Secretary of the ISCA SynSIG (Special Interest Group on Speech Synthesis) and a Co-Organiser of the GENEA Workshops on Embodied Agents' Non-Verbal Behavior. His research focuses on speech synthesis, gesture generation, and multimodal interaction, with contributions to TTS systems, neural networks, and embodied AI.

He leads Digital Futures, a cross-disciplinary research center addressing societal challenges through digital technologies. This center is a collaboration between KTH, Stockholm University, and RISE. His work spans foundational research to industrial applications, emphasizing ethical AI, privacy in voice conversion, and human-robot interaction.

Key research themes include causal reasoning in LLMs, adversarial privacy techniques, and benchmarking frameworks like the GENEA Leaderboard. He has organized international workshops (GENEA 2021-2024) and contributed to standards in TTS evaluation methodologies. His technical innovations include HiFi-Glot for formant synthesis and Matcha-TTS for fast waveform generation.

His research integrates audio, gesture, and motion synthesis with deep learning, addressing challenges in spontaneous speech synthesis, multimodal coherence, and listener perception. He advocates for rigorous evaluation practices and open challenges to advance the field's reproducibility and real-world applicability.