A Visual Question Answering dataset derived from the BD3 Building Defect Dataset. It pairs images of building surfaces with questions and defect category answers, designed for training and evaluating Vision-Language Models. The dataset was created by author 'chandrabhuma' and was last updated on December 27, 2025.
Use Cases
- Training Vision-Language Models for building defect recognition based on paired images and questions.
- Evaluating model performance on visual question answering tasks specific to construction defects.
- Benchmarking multimodal AI systems on domain-specific visual recognition and classification.
Strengths
- Dataset is specifically designed for training and evaluating Vision-Language Models.
- Each sample contains an RGB image paired with a fixed question and a defect category answer.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Last updated 2025-12-27 04:00:33; freshness should be verified.
Provenance
- Source
- Derived from the original BD3 (Building Defect Dataset).
- Freshness
- Last updated 2025-12-27 04:00:33.