3Snet-CLID: High-Fidelity Nanoscale Live-Cell Images for Computational Super-Resolution

A 1.7 GB dataset of images processed by the 3Snet-CLID computational super-resolution method. The method integrates a hybrid supervised/self-supervised deep learning network for denoising with Richardson–Lucy deconvolution. It was created by Fudong Xue and last updated on May 5, 2026.

Use Cases

• Benchmarking computational super-resolution methods based on the described 3Snet-CLID architecture.
• Training denoising models for microscopy images based on the signal-preserving strategy.
• Studying subcellular structures like mitochondrial membranes and nuclear pores based on the described imaging targets.
• Developing accessible platforms for nanoscale live-cell imaging as mentioned in the description.

Strengths

• The underlying method achieves more than 5-fold resolution improvement on conventional microscopes.
• The dataset is licensed under CC-BY-4.0, permitting broad reuse with attribution.
• The method is described as mitigating artifacts and enhancing robustness for diverse structures.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is 1.7 GB, which may require significant storage and bandwidth.

Provenance

Source

figshare

Collection Method

Images were generated using the 3Snet-CLID computational super-resolution method applied to wide-field or spinning-disk confocal images.

Freshness

Last updated 2026-05-05 07:01:05; freshness should be verified.