Datasets and code resources.
This page only uses links that already exist in repository data. If a paper has no verified code or dataset URL, it stays off the list.
Each row points back to the paper page when a local review page exists, and it keeps the external link chips visible for quick reuse.
NasoVoce: A Nose-Mounted Low-Audibility Speech Interface for Always-Available Speech Interaction
No separate code or dataset URL is published in repository data yet.
LipVoicer: Generating Speech from Silent Videos Guided by Lip Reading
Project page and demo material are available for the paper's lip-to-speech setup.
RobustL2S: Speaker-Specific Lip-to-Speech Synthesis exploiting Self-Supervised Representations
Speech samples are published on the project page, with evaluation on Lip2Wav, GRID, and TCD-TIMIT.
WESPER: Zero-shot and Realtime Whisper to Normal Voice Conversion for Whisper-based Speech Interactions
Project page and code links are surfaced in the paper text; the resource package is listed here for quick reuse.
Project page · Code · PDF
SilentSpeller: Towards mobile, hands-free, silent speech text entry using electropalatography
Dataset note: 2328 isolated words per tuned participant, 107 phrases for walking/seated, and a public dataset mention.
SottoVoce: An Ultrasound Imaging-Based Silent Speech Interaction Using Deep Neural Networks
Dataset note: about 500 spoken commands per collaborator across 2 collaborators.
Vid2Speech: Speech Reconstruction from Silent Video
GRID corpus with 1000 videos of speaker S4 for the main reconstruction task, plus cross-speaker and OOV tests.
Speech Synthesis from Silent Video: A Deep Learning Approach
Project resources support speech reconstruction from silent video and point to the original code/sample page.