Loading page…
Loading page…
A completed 12-second, 480p Hotel Lobby AI video using two portrait references, with restored source audio.
The model replaces both performers from 0:00 of the source. After generation, the original track is restored so the result keeps the song and timing of this segment.
Sample output. This is one supplied completed generation, separate from the studio demonstration and your private videos. Compare both faces with the portrait references below. Watch the separate AI demonstration in the Hotel Lobby studio.
Left · Image 1Assigned to the person on the left of the source. One clear adult face in a separate photo.
Right · Image 2Assigned to the person on the right. Swap sides in the studio if the order is wrong before you generate.
Both people share the same shot and react within one original performance. Separate portrait references are assigned to the left and right performer, so the two identities can be reviewed individually.
The studio replaces both performers in a prepared video using your separate portrait references. The source supplies the timing and camera movement. This two-role workflow is explained on the Hotel Lobby character swap page. The workflow then restores the source audio rather than generating a new song or voice.
Read the Hotel Lobby AI creation guide for photo selection, demonstration playback and output checks.
Your private photos and results are not used as a public customer gallery. See our media privacy policy.