Yuanlong Zhang from Tsinghua University created this dataset for a deep learning project on large depth-of-field ultra-compact microscopes. It includes blurry test images, sharp network output images, and pre-trained network weights. The dataset is hosted on Papers with Code and is licensed as Open Access.
Use Cases
- Benchmarking image deblurring algorithms based on the provided blurry and sharp image pairs.
- Fine-tuning pre-trained neural networks for other microscopy applications based on the included network weights.
- Developing new computational imaging pipelines based on the progressive optimization method described.
- Training models for depth-of-field extension in compact optical systems based on the dataset's domain.
Strengths
- Includes pre-trained network weights, which can accelerate model development.
- Provides paired blurry and sharp images, which are essential for supervised learning tasks.
- Originates from a research project at Tsinghua University, suggesting academic rigor.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Row count and file size are unknown, which may limit suitability assessment.
- Column-level documentation is absent; field semantics must be inferred after download.
Provenance
- Source
- Yuanlong Zhang, Tsinghua University
- Collection Method
- Likely generated as part of a research project on computational microscopy.