Identity-Preserving Text-to-Video Generation by Frequency Decomposition

Benchmark Model Rank Results
open-domain-subject-to-video-on-opens2v-evalKling 1.6#1Total Score: 0.5446Venue: Close-SourceAesthetics: 0.4460
open-domain-subject-to-video-on-opens2v-evalPika 2.1#6Total Score: 0.4888Venue: Close-SourceAesthetics: 0.4687
open-domain-subject-to-video-on-opens2v-evalVidu 2.0#8Total Score: 0.4759Venue: Close-SourceAesthetics: 0.4147