FluoGen: Preprocessed Fluorescence Microscopy Images from 20 Public Datasets

488.9 MB of preprocessed fluorescence microscopy images aggregated from 20 public datasets including 2018 Data Science Bowl, BBBC, BioSR, and RxRx. This dataset, created by HuaiAn Chen, serves as a training resource for the FluoGen project and was last updated on May 16, 2026. The data retains the original licenses of its constituent sources.

Use Cases

• Train cell segmentation models based on fluorescence microscopy images from datasets like BBBC and Fluo-N2DL-HeLa.
• Benchmark image-to-image translation algorithms using paired microscopy data from sources like BioSR.
• Develop and evaluate tracking models for dynamic cellular processes using time-series data from DynamicNuclearNet-tracking-v1_0.

Strengths

• Aggregates data from 20 distinct public sources, providing a broad base of biological imaging scenarios.
• Preprocessed for consistency, likely reducing initial data preparation effort for model training.
• The 488.9 MB size suggests a medium-scale collection suitable for many deep learning experiments.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific model architectures.
• Data may reflect bias inherent to the specific biological experiments and imaging conditions of the aggregated sources.

Provenance

Source

Aggregated from 20 public datasets listed in the description (e.g., 2018 Data Science Bowl, BBBC, RxRx).

Collection Method

Preprocessed and organized from publicly available datasets for the FluoGen project.

Freshness

Last updated 2026-05-16 12:18:31; freshness should be verified.