Samsung Doc Classifier: Movie Posters and Magazine Covers in Phone Scenes

Movie posters and magazine covers composited into realistic phone-like scenes. The dataset appears designed for computer vision tasks involving document classification within a synthetic environment. Its author, organization, and specific scale are unknown.

Use Cases

• Train a classifier to distinguish movie posters from magazine covers based on visual features.
• Develop object detection models to locate documents within a composite phone-screen scene.
• Benchmark model robustness on synthetic data that mimics real-world mobile camera captures.
• Test image segmentation algorithms on documents placed within a complex background.

Strengths

• Contains two distinct document types: movie posters and magazine covers.
• Images are composited into realistic scenes, which may provide useful context for training.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Collection Method

Likely involves compositing and image processing to create synthetic scenes.