{"id":29224588,"url":"https://github.com/sing-group/polydeep-object-detection","last_synced_at":"2025-07-03T06:07:25.495Z","repository":{"id":47192615,"uuid":"394380991","full_name":"sing-group/polydeep-object-detection","owner":"sing-group","description":null,"archived":false,"fork":false,"pushed_at":"2022-04-13T18:30:14.000Z","size":292,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":6,"default_branch":"master","last_synced_at":"2023-02-27T07:56:26.288Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/sing-group.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2021-08-09T17:25:34.000Z","updated_at":"2022-07-20T10:55:18.000Z","dependencies_parsed_at":"2022-09-10T13:22:34.585Z","dependency_job_id":null,"html_url":"https://github.com/sing-group/polydeep-object-detection","commit_stats":null,"previous_names":[],"tags_count":null,"template":null,"template_full_name":null,"purl":"pkg:github/sing-group/polydeep-object-detection","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sing-group%2Fpolydeep-object-detection","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sing-group%2Fpolydeep-object-detection/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sing-group%2Fpolydeep-object-detection/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sing-group%2Fpolydeep-object-detection/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sing-group","download_url":"https://codeload.github.com/sing-group/polydeep-object-detection/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sing-group%2Fpolydeep-object-detection/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":263271501,"owners_count":23440396,"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":[],"created_at":"2025-07-03T06:07:23.985Z","updated_at":"2025-07-03T06:07:25.462Z","avatar_url":"https://github.com/sing-group.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# PolyDeep project\nThis project is part of the PolyDeep research project. See more details at http://polydeep.org\n\n# Compi pipelines\n\nFour different Compi pipelines were developed for the object-detection model development and evaluation: \n- (i) dataset split for development (training and validation datasets) and testing\n- (ii) model development\n- (iii) model evaluation (with images)\n- (iv) video annotation. \n\n## Dataset split pipeline (`ttv.xml`)\n\nThis pipeline is a preamble pipeline containing the steps necessary to prepare the dataset for the subsequent model development and evaluation. This pipeline starts with the check-dataset step, which is a control step that checks whether the dataset has been downloaded locally and, if not, alerts the user and ends the execution. Then, the create-ttv step analyses the dataset to generate training, validation, and test subsets based on the number of images of each polyp (see Sect. 4.1). As a result, this step generates three files with the polyp identifiers that correspond to each subset of the data.\n\n## Model development pipeline (`train.xml`)\n\nThis pipeline is responsible for the model development. This pipeline also starts with the check-dataset step, which is followed by another control step (check-gpu) that checks the sanity of the GPU memory, as a failure in it could cause erroneous results. After that, the model-development step performs the model training and validation, generating several files with the performance results. This step is followed by the generate-plot-data that post-processes previous result files in order to plot them in the three following steps: plot-loss, plot-map, and plot-metrics, which generate several plots that summarize the model evaluation results. Finally, a cleanup step removes some auxiliary files generated during the pipeline execution.\n\n## Model evaluation pipeline (`test.xml`) \n\nThis pipeline is responsible for evaluating the model selected in the development phase with a new/unseen dataset of images. This pipeline has the same control steps as the model development pipeline (i.e. check-dataset and check-gpu) and a new control step (check-neural-network) to check that the trained model exists. Then, the test step evaluates the model with the test dataset created with the dataset split pipeline, generating a file with the performance results.\n\n## Video annotation pipeline (`video-annotation.xml`)\n\nThis pipeline allows applying the model to video segments of polyp and normal-mucosa regions to obtain annotated videos. This pipeline starts with the same control steps that the model evaluation pipeline. Then, this pipeline comprises three main stages: (i) download videos (download-polyp-videos and download-normal-mucosa-videos), (ii) extract video segments (extract-polyp-segments and extract-normal-mucosa-segments), and (iii) use the model to detect polyps in each frame of the video segments (predict-polyp-segments and predict-normal-mucosa-segments). A final a cleanup step removes some auxiliary files generated during the pipeline execution.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsing-group%2Fpolydeep-object-detection","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsing-group%2Fpolydeep-object-detection","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsing-group%2Fpolydeep-object-detection/lists"}