{"id":21115900,"url":"https://github.com/thomasjo/nemo","last_synced_at":"2025-03-14T10:26:59.699Z","repository":{"id":49305012,"uuid":"176238839","full_name":"thomasjo/nemo","owner":"thomasjo","description":"Detection and classification of microscopic foraminifera","archived":false,"fork":false,"pushed_at":"2022-12-08T03:37:51.000Z","size":419,"stargazers_count":3,"open_issues_count":5,"forks_count":0,"subscribers_count":6,"default_branch":"master","last_synced_at":"2025-01-21T04:41:40.038Z","etag":null,"topics":["climatology","deep-learning","foraminifera","geology","geoscience","machine-learning","oceanography","research","tensorflow"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/thomasjo.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2019-03-18T08:39:57.000Z","updated_at":"2022-01-25T16:36:14.000Z","dependencies_parsed_at":"2023-01-25T04:15:58.218Z","dependency_job_id":null,"html_url":"https://github.com/thomasjo/nemo","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/thomasjo%2Fnemo","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/thomasjo%2Fnemo/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/thomasjo%2Fnemo/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/thomasjo%2Fnemo/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/thomasjo","download_url":"https://codeload.github.com/thomasjo/nemo/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243560848,"owners_count":20310993,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["climatology","deep-learning","foraminifera","geology","geoscience","machine-learning","oceanography","research","tensorflow"],"created_at":"2024-11-20T02:01:54.415Z","updated_at":"2025-03-14T10:26:59.680Z","avatar_url":"https://github.com/thomasjo.png","language":"Python","readme":"# Project Nemo\n\nForaminifera (forams for short) classification via deep feature extraction.\n\n## Image dataset\n\nAll models have been trained on a dataset of large, high-resolution images of\nforams. The dataset has been produced by our research group, and will be made\npublically available in the near future. Each of the source images consist of\na single class of forams. From these images, patches of _224x224_ pixels are\nextracted using combinations of Gaussian smoothing, binary image generation\nvia thresholding, and connected components. The first step removes the metallic\nborder present in all source images, and the second step extracts candidate\npatches. Each patch that passes a defined selection critera is extracted by\nplacing a _224x224_ crop at the centroid of the candidate region. The entire\nprocess is automated in the `preprocess_data.py` script.\n\nOnce the source images have been preprocessed by extracting patches, datasets\nfor training, validation, and testing are generated automatically by using the\n`build_datasets.py` script.\n\n### Caveat regarding `raw-halves` source images\n\nThe `raw-halves` source images are slightly different in nature, and requires\nthat the `preprocess_data.py` script be invoked with `--border-threshold=50`.\nPatches from this dataset must be manually copied to the `preprocessed` folder\nbuilt by process outlined above. In the future, we should find a way to fully\nautomate this step as well.\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fthomasjo%2Fnemo","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fthomasjo%2Fnemo","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fthomasjo%2Fnemo/lists"}