41605 People - Multiple Style Video Data, which includes multiple styles video data of 41605 person IDs in various environment. The person IDs in this dataset covers different skin colors like white/yellow/brown/black and different ages like young/middle-age/old-age. The resolution and duration of each video is no less than 1080p and 10s, and all videos have audio. This dataset can be used for character consistent video generation and digital human generation tasks.

For more details, please refer to the link: https://www.nexdata.ai/datasets/llm/1689?source=Opendatabay

41605 people, each with 1 frontal photo and 4 to 50 videos, where the videos include at least 3 of the following 4 types: ①videos of person’s talking. ②videos of person’s body motion. ③videos of person-object-interaction. ④videos of person-person-interaction.

①all videos have audio. ②video’s resolution is no less than 1920x1080, and the duration is no less than 10 seconds. ③the environment, background, and clothing of each person vary across the videos. ④the facial information in the videos is complete.

skin colors cover white/yellow/brown/black, genders cover female/male, and ages cover young/middle-age/old-age, etc.

skin color, gender, and ages of each person, as well as the type corresponding to each video.

Commercial License