Experiment Timeline Dashboard
Machine Learning Model Training Experiments Across Biological Groups
Machine Learning Model Training Experiments Across Biological Groups
Tracking and visualizing the progress of machine learning experiments designed to automatically segment cellular organelles from high-resolution electron microscopy data. Our mission is to advance automated image analysis for biological discovery through systematic model training and evaluation.
Explore comprehensive visualizations of our machine learning experiment timeline, performance metrics, and training progress across different biological datasets.
Explore the chronological progression of all experiments with detailed hover information, group-based color coding, and real-time status updates.
View TimelineVisualize experiment durations, overlaps, and training periods in a comprehensive timeline format showing real training dates and progress.
View Gantt ChartComprehensive statistics and breakdowns by experiment group, model type, organelle targets, and training configurations.
View StatisticsAnalyze dataset and crop usage across all experiments, identify most frequently used datasets, and track data distribution.
View Dataset StatsClear, easy-to-read bar charts showing F1 scores for each organelle. Separate pages for each organelle type with detailed rankings.
View ScoresQuick reference table showing the top F1 score achieved for each organelle with setup and iteration information.
View Best ScoresTrack how F1 scores improve across training iterations. See which setups learn faster and when performance plateaus.
View ProgressionAccess the complete experiment metadata in CSV format for custom analysis, including training parameters and performance metrics.
Download CSVDetailed crop-level dataset usage information in YAML format, showing exact crops used per experiment.
Download YAMLExplore our interactive data portal with high-resolution EM images and annotations used for training these models.
Visit PortalDetailed technical documentation about experiment configurations, model architectures, and training methodologies.
Read Documentation