Face likeness

How 27 models handled “Face likeness” from our eval suite, Image to Video (Kling V3 Standard) first, then by image-to-video rating.

What we look for

The person shown at the end is recognizably the same as in the input image. Temporally stable through the bus pass.

Prompt

“A man stands in an asian food alley way holding a bowl of ramen. The chopsticks fall on the ground and a bus passes in front of him. At the end show the same person”

Input

Tests for

Consistency, reliability