Diffuman4D Example: Test Scenes for 4D Human View Synthesis

Several example scenes for testing the Diffuman4D model, which synthesizes 4D consistent human views from sparse-view videos. The dataset was uploaded by author 'krahets' to Hugging Face and last updated on September 11, 2025. It is described as test data not part of the official DNA-Rendering dataset.

Use Cases

• Testing 4D consistent human view synthesis models based on the described sparse-view video input.
• Benchmarking spatio-temporal diffusion models for human-centric scene generation.
• Evaluating the performance of the Diffuman4D method on provided example scenes.

Strengths

• Last updated on 2025-09-11, indicating recent maintenance.
• Linked to a specific research paper and model (Diffuman4D), providing clear context.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count, file formats, and column-level documentation are absent.
• The data is explicitly described as example test scenes, not part of a formal training dataset.

Provenance

Source

Hugging Face dataset uploaded by user 'krahets', associated with the Diffuman4D research project.

Collection Method

Likely generated or curated as example test data for the associated model.

Freshness

Last updated 2025-09-11 11:14:54.