MET_S01: Scene-Level Videos for AI-Generated Film Style Analysis

Scene-level video materials from the 1927 film 'Metropolis' by Fritz Lang, used for comparative analysis of original and AI-generated content. The 70.8 MB collection includes RAW and PRE-processed MP4 videos, prompt files, generation settings, and metadata records. Jinyu Kim created this dataset for a doctoral dissertation on film style reconfiguration via generative AI.

Use Cases

• Compare original and AI-generated video styles based on scene-level materials.
• Analyze the effects of generation settings and prompts on video output based on included metadata.
• Study the reconfiguration of German Expressionist film style through generative models based on the described film context.

Strengths

• Includes both RAW generated videos and standardized PRE-processed videos at 720p, 24 fps, and 8 seconds.
• Provides prompt files and generation settings, offering transparency into the AI video creation process.
• Clearly defined scene-level organization and metadata for the specific film 'Metropolis' (1927).

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Audio is excluded from the analysis, limiting multimodal study.

Provenance

Source

figshare

Collection Method

Part of a doctoral dissertation project; likely created via generative AI video models.

Time Range

Source film released in 1927.

Freshness

Last updated 2026-06-04 21:01:09; freshness should be verified.