Jaeyoung Huh's dataset, last updated in April 2026, contains comparison results of FID scores. The data originates from a study using a switchable CycleGAN to enhance liver ultrasound images from an older device, with images from a newer device as targets. The dataset is 5.5 KB in size and is available in XLS format.
Use Cases
- Benchmarking image enhancement algorithms based on reported FID score comparisons.
- Training models for medical image quality improvement based on the described unsupervised learning setup.
- Analyzing the correlation between algorithmic image enhancement and clinical diagnostic performance mentioned in the study.
Strengths
- Dataset is openly licensed under CC-BY-4.0.
- Specific evaluation metrics (FID scores) are provided for a defined medical imaging task.
- The underlying study reported statistically significant improvements in image brightness, contrast, and overall quality (p < 0.001).
Limitations
- Row count is unknown, which may limit suitability assessment.
- Column-level documentation is absent; field semantics must be inferred after download.
- The dataset is very small at 5.5 KB, indicating limited scope.
Provenance
- Source
- figshare, author Jaeyoung Huh.
- Collection Method
- Likely derived from a study comparing FID scores for enhanced liver ultrasound images generated by a deep learning algorithm.
- Time Range
- null
- Freshness
- Last updated 2026-04-28 17:44:29; freshness should be verified.
- Geography
- null