Open Source
Audiobook-CC: Controllable Long-context Speech Generation for Multicast Audiobook
arXiv eess.AS0 views
arXiv:2509.17516v2 Announce Type: replace Abstract: Existing text-to-speech systems predominantly focus on single-sentence synthesis and lack adequate contextual modeling as well as fine-grained performance control capabilities for generating coherent multicast audiobooks. To address these limitations, we propose a context-aware and emotion controllable speech synthesis framework specifically engineered for multicast audiobooks with three key innovations: a context mechanism for contextual consistency, a disentanglement paradigm to decouple style control from speech prompts for semantic consistency, and self-distillation to boost emotional expressiveness and instruction controllability. Experimental results show superior performance across the generation of narration, dialogue, and the whole chapter, significantly outperforming existing baselines. Ablation studies are conducted to validate the effectiveness of our proposed methods. Demo samples can be found in https://semisemi-ux.github.io/
Source:https://arxiv.org/abs/2509.17516PreviousSex Differences in the Detection of Parkinson's Disease from SpeechNextInterference-Aware Tone Selection for Bluetooth Channel Sounding
Related posts
Аудит прав доступу
0dev.to WebDev
Show HN: SSH OS – an SSH-only operating system for the 1987 Macintosh SE
0Hacker News Show
Show HN: Citeparse: parse reference lists and flag invented citations on a CPU
0Hacker News Show
Show HN: Ufatfs: Unix overlay file system for VFAT/exFAT
0Hacker News Show
The Address Layer — Or, "The Thing Moved" Is Not the Same as "The Old Address Is Dead"
0dev.to
Making grill-me ask multiple-choice questions in Claude Code
0dev.to