Handvqa

HandVQA is a dataset introduced in a CVPR 2026 paper by researchers from UNIST, University of Aberdeen, University College London, and Fogsphere. It is designed for diagnosing and improving fine-grained spatial reasoning about hands in vision-language models. The dataset page is hosted on Hugging Face by author kcsayem and was last updated on March 30, 2026.

Use Cases

• Benchmarking model performance on fine-grained spatial reasoning tasks based on the dataset's described purpose.
• Diagnosing specific weaknesses in vision-language models' understanding of hand-related spatial concepts.
• Training or fine-tuning models to improve performance on visual question answering involving hand poses and interactions.

Strengths

• Dataset is associated with a peer-reviewed CVPR 2026 conference paper, suggesting academic rigor.
• Focus is on a specific, challenging task: fine-grained spatial reasoning about hands.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.
• License information is unknown, which could restrict usage.

Provenance

Source

Hugging Face dataset uploaded by kcsayem, associated with academic institutions.

Collection Method

Likely created for research purposes as described in the CVPR 2026 paper; specific collection method is not detailed in the provided input.

Freshness

Last updated 2026-03-30 17:45:59; freshness should be verified.