TARDIS-em v0.3.0

• Small code restrictions
• Update scaling
• fixes in filament analysis and added new analysis
• Added full support for Filament/Microtubule prediction from TIRF fluorescent images
• Updates in the Napari plugin for Filament/Microtubule prediction from TIRF
• New workflow in Napari to predict filaments and perform corrections/analysis

Full Changelog: v0.2.8...v0.2.9

TARDIS-em v0.2.9

• Fixed bugs and optimized code from v0.2.8

Full Changelog: v0.2.8...v0.2.9

TARDIS-em v0.2.8

• Fixed numpy v2.0.0 support
• Added general predictor for filament and object-type structures
• Added support for model versioning
• Users are now allowed to use starting from v0.2.8 new and old model version
• Added prediction metadata to all save files, including the prediction log file
• Fixes from v0.2.6 version
• Change scaling types for up- and down-scaling of images before/after predictions
• Added adaptive threshold as an optional cnn threshold
• Added model for actin

TARIDS-em v0.2.6

What's new:

Full Changelog: v0.2.5...v0.2.6

• General changes:
  • General:
    • Bugfix when multiple files are predicted in batch

TARIDS-em v0.2.5

What's new:

Full Changelog: v0.2.4...v0.2.5

• General changes:
  • General:
    • Added support for predicting Actin
    • Predicting point clouds directly from cli

TARIDS-em v0.2.4

Full Changelog: v0.2.2...v0.2.4

What's new:

This intends to be a release submitted with Nature Method 2024

• General changes:
  • General:
    • Few fixes from v0.2.2
    • Added visualization for semantic masks
    • Documentation
    • Fix conda upload

TARDIS-em v0.2.2

Full Changelog: v0.2.1...v0.2.2

TARDIS-em v0.2.1

Full Changelog: v0.1.1...v0.2.1

What's new:

This intends to be a release submitted with Nature Method 2024

• General changes:

  • General:
    • Improve prediction for microtubules and membranes (reduce false positive)
    • Update Membrane and Microtubule modules predictions
    • Update usage tutorials
    • Added pypi and conda installations
    • Enabled scripting with tardis-em

• SpindleTorch module changes:

  • General:
    • Update Fnet_attn model

• DIST module changes:

  • Optimize:
    • Re-trained DIST model using simulated datasets
    • Build 2 model for:
      • filaments and general 2D structures
      • 3D objects like membranes mitochondria LiDAR data etc.

TARDIS-em v0.1.1

What's new:

This intends to be a release submitted with Nature Method 2023

• General:
  • General:
    • Documentation update
    • Bug-fixes
    • General improvement in predictions

Full Changelog: v0.1.0...v0.1.1

TARDIS-em v0.1.0

What's new:

This intends to be a release submitted with Nature Method 2023

• General:

  • General:
    • Documentation update
    • Added full support for OTA updates of the entire package
    • Fixed AWS access denied error on some networks
    • A few bug fixes
    • Fixed Bugs in final filament filtering algorithms
    • Added filament filtering for removing false-positive rapid 150-degree connections
    • Microtubule output is now sorted by the length
    • Each instance receives a segmentation confidence score by which the user can filter out predictions

• New_Feature:

  • Added new output format .ply
  • New general tardis call
  • Added helper functions csv_am and am_csv
  • Added instance prediction from semantic binary masks

• Optimize:

  • Added an optional checkpoint to all Tardis calls
  • Improvements in training for CNN and DIST by users
  • Amira possible output as a raw point cloud

• BugFix:

  • Fixed save for .mrc files

• SpindleTorch module changes:

• General:

  • Retrained FNet_32 model for membrane and microtubules
  • Train FNet_32 for 2D membrane segmentation

• Optimize:

  • 2D CNN network set-up

• DIST module changes:

• General:

  • Added simulated data for training on filament-like structures
  • Re-train model no simulated + real data
  • Fine-tuned setting for predictions and post-processing

• New_Feature:

  • Experimental SparseDist model to offer more memory-efficient performance,
    for instance segmentation

• Optimize:

  • Improved visualization outputs
  • Mcov metric optimization
  • Rebuild Graph prediction function to be more robust
  • Reverse-engineered Open3D voxal downsampling and added random downsampling
  • Added distance embedding with a range value

What's Changed

• pull by @RRobert92 in #74
• Merge pull request #74 from SMLC-NYSBC/main by @RRobert92 in #75

Full Changelog: v0.1.0RC2-hotfix2...v0.1.0RC3

What's Changed

• Update sphinx_documentation.yml by @RRobert92 in #76
• Update sphinx_documentation.yml by @RRobert92 in #77
• Update README.md by @RRobert92 in #79

Full Changelog: v0.1.0RC3...v0.1.0