Bibliographic record
Abstract
Summaries of files:- Python scripts (.py): Core implementation of data processing, training, inference, evaluation, and visualization.- Jupyter Notebooks (.ipynb): Interactive notebooks for data exploration and visualizations.- Markdown files (.md): Documentation for the repository and specific sub-modules.- Configuration and Metadata files: JSON (.json) for dataset definitions/interconnectedness, CITATION.cff for citations, and LICENSE.Note: The repository primarily contains the software components. Expected data formats manipulated by these scripts include HDF5 (.h5), NIfTI (.nii.gz), and image sequences (.png/.jpg). Summaries of tools or codes:- Python 3.x is required to execute the scripts and notebooks.- Deep Learning Frameworks: PyTorch and MONAI (Medical Open Network for AI) are required for the training and inference of the 3D models (Swin-UNETR, SegResNet, UNet-3D).- META's Segment Anything Model 2 (SAM 2) is required for the scripts in the sam2 directory.- Data Processing & Visualization Libraries: h5py (HDF5 data parsing), nibabel (NIfTI formatting), numpy, matplotlib, k3d (3D visualizations), and standard scientific Python tools. Explanations of the uses for uploaded scripts or codes:- analysis/: Lightweight analysis scripts for exploratory checks, such as volume energy plotting.- data/: Conversion tools to translate simulation outputs into deep learning compatible formats (HDF5 to projections/slices, PNG to JPG, image to NIfTI) and dataset JSON generation.- evaluation/: Scripts (e.g., evaluate-nii.py) for metric computations to evaluate model segmentation outputs against ground truth masks.- graphs/: Utility scripts and notebooks designed to generate quantitative charts, histograms, and figures for analysis.- post-process/: Scripts to refine predictions by cleaning data, merging masks, removing unwanted pieces, converting semantic to instance segmentation, and multiple 3D visualizers (mag3d, temp3d, vel3d).- sam2/: Scripts exploring superbubble instance tracking and video inference using point-prompt SAM 2.- segresnet/, swin-unetr/, unet-3d/: Directories containing the primary training, testing, and validation routines for the respective 3D deep learning architectures deployed on the simulation data. Summary of the relationship between the files uploaded and the related article:This uploaded repository contains the complete software pipeline designed and developed in support of the submitted article "Segmenting Superbubbles in a simulated Multiphase Interstellar Medium using Computer Vision". The codes encompass every stage of the computational methodology discussed in the paper: from transforming raw multiphase ISM simulation data into suitable ML formats, training the state-of-the-art vision models (like Swin-UNETR and SAM2) to detect and segment superbubbles, post-processing the output masks, to finally evaluating the results and producing the quantitative tables and visual graphs presented in the article.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.018 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.003 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.007 |
| Open science | 0.006 | 0.005 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.431 | 0.516 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".