ExoDeepFinder v0.3.14

ExoDeepFinder provides a GUI which enables to call each of the ExoDeepFinder commands:

edf_convert_tiff_to_h5              # convert tiff folders to a single h5 file
edf_segment                         # segment a movie
edf_generate_annotation             # generate an annotation file from a segmentation by clustering it
edf_generate_segmentation           # generate a segmentation from an annotation file
detect_spots_with_atlas         # detect bright spots in movies with the Atlas detector
detect_spots                    # detect bright spots in movies (with any detector)
edf_merge_detector_expert           # merge the expert annotations with the detector segmentations for training
edf_structure_training_dataset      # structure dataset files for training
edf_train                           # train a new model
exodeepfinder                       # combine all above commands

Just uncompress and launch exodeepfinder.app (on Mac), exodeepfinder.exe (on Windows) or run exodeepfinder from a terminal (on Linux).

---

Changes: This new version better integrates with Atlas, the bright spot detector.

ExoDeepFinder v0.3.6

ExoDeepFinder provides a GUI which enables to call each of the ExoDeepFinder commands:

edf_convert_tiff_to_h5              # convert tiff folders to a single h5 file
edf_segment                         # segment a movie
edf_generate_annotation             # generate an annotation file from a segmentation by clustering it
edf_generate_segmentation           # generate a segmentation from an annotation file
edf_detect_spots                    # detect bright spots in movies
edf_merge_detector_expert           # merge the expert annotations with the detector segmentations for training
edf_structure_training_dataset      # structure dataset files for training
edf_train                           # train a new model
exodeepfinder                       # combine all above commands

Just uncompress and launch exodeepfinder.app (on Mac), exodeepfinder.exe (on Windows) or run exodeepfinder from a terminal (on Linux).

---

Changes: This new version trains using curriculum learning: a first model is trained on small patches, then its weights are used to train a model with bigger patches, and this process continues until reaching a proper patch size.

ExoDeepFinder v0.3.3 with Tensorflow 2.11.1

ExoDeepFinder provides a GUI which enables to call each of the ExoDeepFinder commands:

edf_convert_tiff_to_h5              # convert tiff folders to a single h5 file
edf_segment                         # segment a movie
edf_generate_annotation             # generate an annotation file from a segmentation by clustering it
edf_generate_segmentation           # generate a segmentation from an annotation file
edf_detect_spots                    # detect bright spots in movies
edf_merge_detector_expert           # merge the expert annotations with the detector segmentations for training
edf_structure_training_dataset      # structure dataset files for training
edf_train                           # train a new model
exodeepfinder                       # combine all above commands

Just uncompress and launch exodeepfinder.app (on Mac), exodeepfinder.exe (on Windows) or run exodeepfinder from a terminal (on Linux).

ExoDeepFinder 0.2.3

The first release of ExoDeepFinder provides a GUI which enables to call each of the ExoDeepFinder commands:

edf_convert_tiff_to_h5              # convert tiff folders to a single h5 file
edf_segment                         # segment a movie
edf_generate_annotation             # generate an annotation file from a segmentation by clustering it
edf_generate_segmentation           # generate a segmentation from an annotation file
edf_detect_spots                    # detect bright spots in movies
edf_merge_detector_expert           # merge the expert annotations with the detector segmentations for training
edf_structure_training_dataset      # structure dataset files for training
edf_train                           # train a new model
exodeepfinder                       # combine all above commands

Just uncompress and launch exodeepfinder.app (on Mac), exodeepfinder.exe (on Windows) or run exodeepfinder from a terminal (on Linux).

The Linux release is big (over 4Gb) because it contains the libraries required for the GPU acceleration. Thus they are split in two parts (ExoDeepFinder_Linux-x86_64_part1.tar.gz and ExoDeepFinder_Linux-x86_64_part2.tar.gz). To uncompress them, use the following command: tarcat ExoDeepFinder_Linux-x86_64_part*.tar.gz | tar -xvzf -.