StoryReasoning: 4,178 Cohesive Visual Stories from Movie Scenes

The StoryReasoning dataset contains 4,178 cohesive visual stories derived from 52,016 images. It organizes temporally connected image sequences extracted from the same movie scenes to ensure narrative coherence. The dataset was created by author daniel3303 and was last updated on the Hugging Face platform in May 2025.

Use Cases

• Training models for consistent character identity across multiple images based on the cross-frame consistency feature.
• Evaluating narrative coherence in generated visual stories based on temporally connected image sequences.
• Benchmarking multi-image story generation systems using the dataset's 4,178 cohesive stories.
• Researching visual narrative structure using sequences extracted from movie scenes.

Strengths

• Contains 4,178 cohesive stories, providing a substantial collection for training and evaluation.
• Derived from 52,016 images, offering a large visual corpus.
• Sequences are extracted from the same movie scenes, which likely ensures narrative coherence.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset's source and potential biases inherent to movie scenes are not detailed.

Provenance

Source

huggingface

Collection Method

Temporally connected image sequences extracted from movie scenes.

Freshness

Last updated 2025-05-16 12:54:12; freshness should be verified.