Xinlei Niu PhD Presentation: Generative Audio

• Date: Thursday 25 September, 2025
• Time: 10:00am
• Venue: 4.15 Board Room, Level 4, Hanna Neumann Building 145.
• Presenter: Xinlei Niu

PhD researcher Xinlei Niu will deliver her Oral Presentation on Thursday 25 September 2025 at 10:00am in 4.15 Board Room, Level 4, Hanna Neumann Building 145. All welcome!

Title:

Generative Audio Synthesis in Sound, Speech, and Music

Abstract:

Audio synthesis plays an important role in entertainment and creative industries, supporting applications such as Foley sound design, music editing, and speech generation. Traditional methods rely on manual processes, such as hand-crafted Foley, audio remixing, or recording voice actors, that are labor-intensive and demand specialised expertise. Recent advances in generative modelling have transformed this landscape, enabling expert-level results from simple prompts and multimodal inputs. By lowering the barriers to performing complex audio synthesis, these models make advanced sound, speech, and music generation more accessible to a wider range of users. This thesis explores generative audio synthesis in sound, speech, and music, with the goal of enhancing the quality, versatility, and creative potential of audio technologies across diverse tasks.