Use an authorized personal video avatar to replace the real camera with low latency to attend online meetings
Users create a personal avatar by submitting a selfie video, select the virtual camera provided by this site in the conference software, and drive the avatar's screen to speak synchronously when they speak; at the same time, they obtain auditable records of material authorization, compatibility and delay quality.
- 01Materials and LicensingUpload your own selfie video and record authorization, storage period, deletion entry and manual approval status.
- 02Virtual camera previewStart the virtual camera and confirm that the device is enumerable, can output video tracks, and is in allowed conferencing mode.
- 03Meeting readiness testVerify audio driver, lip sync, latency target, and target platform compatibility status.
- 04Publishing and auditingKeep manual approvals, failure records, deletion records, and test reports to avoid treating future assumptions as current functionality.
Meeting Avatar Readiness Report
Results, sources, failures, approvals, and rollbacks are saved by project.
No business results yet
Please first upload a self-portrait video of about 5 minutes authorized by me.
Data use boundary
Selfie videos, facial avatars, authorization instructions, test audio clips and latency test logs are sensitive data. Before uploading, the purpose, storage period, deletion entry, and whether to superimpose a disclosure watermark must be clarified; if the consent_id or authorization record is missing, the system will refuse to process the material. Current 10 role review supplements: The scope of sensitive data must include selfie videos, authorization records, generated/intermediate clone materials, recoverable face frames, virtual camera output samples, test audio, quality test logs, approval notes, and deletion/withdrawal records. ; Selfie videos and test audios only retain the minimum period required to complete the build or test by default; test reports only retain indicators and failure reasons by default. ; After deletion, subsequent generation, preview, testing, export and rollback must not read deleted sensitive materials; the backup/cache cleaning window must be clear. ; usesWeb shall not upload selfie videos, audios, avatar materials, authorized texts or conference content to third parties. ; Exports with default desensitization and does not include raw video, test audio, recoverable face frames, full model raw responses, keys, or cross-workspace data.
Data retention
By default, it is saved until the user deletes it or the project is closed; after deletion, the material recording status changes to deleted, and the original video cannot be read by the generation task and virtual camera session. Test reports can retain indicators and failure reasons, but should not continue to retain recoverable face videos unless otherwise authorized by the user.
Human responsibility and rollback
This product will process user selfie videos, facial images and possible microphone test audio, and is only used to generate and test video conferencing clones authorized by the user. AI avatars may cause misidentification or abuse. Users should abide by conference rules, platform policies and conference disclosure requirements; this site does not provide the ability to automatically answer questions, speak beyond authority, or impersonate others.
Rollback only refers to undoing the avatar material status, virtual camera session status, test report release status or workflow stage in this site; this product is not connected to the external conference platform production system, so rollback will not and cannot modify external conferences, accounts, meeting records or platform settings.