{"id":17606418,"url":"https://github.com/atcold/torch-inria","last_synced_at":"2025-04-30T12:20:28.131Z","repository":{"id":19285909,"uuid":"22522858","full_name":"Atcold/torch-INRIA","owner":"Atcold","description":"Loads face, torso, body and background samples from INRIA dataset","archived":false,"fork":false,"pushed_at":"2014-08-01T22:33:17.000Z","size":444,"stargazers_count":6,"open_issues_count":1,"forks_count":2,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-03-30T16:05:49.900Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Lua","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/Atcold.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":"2014-08-01T18:36:48.000Z","updated_at":"2025-03-02T15:41:47.000Z","dependencies_parsed_at":"2022-09-13T17:51:15.065Z","dependency_job_id":null,"html_url":"https://github.com/Atcold/torch-INRIA","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/Atcold%2Ftorch-INRIA","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Atcold%2Ftorch-INRIA/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Atcold%2Ftorch-INRIA/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Atcold%2Ftorch-INRIA/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Atcold","download_url":"https://codeload.github.com/Atcold/torch-INRIA/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":251700432,"owners_count":21629837,"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":"2024-10-22T15:44:26.718Z","updated_at":"2025-04-30T12:20:28.072Z","avatar_url":"https://github.com/Atcold.png","language":"Lua","readme":"# torch-INRIA\n\n[`makeClasses`](src/makeClasses.lua) is a simple script that loads face, torso, body and background samples from INRIA dataset and saves them in a *ImageNet* compatible format. This means, object are fairly centred and the sample size is 256 × 256 px².\n\nIn this first release, the dataset has to be downloaded manually from [here](http://pascal.inrialpes.fr/data/human/INRIAPerson.tar) and extracted into `~/Work/Datasets`.\n\n\u003e In the code there is my user name (`~` = `/Users/atcold`). I will probably update it in a following commit.\n\nWhat this script does is simply extract every *face*, *torso* and *body* from the positive and *background* from the negative folder, cropping them square and saving them into a `Extracted-data` folder, under a sub-directory representing their label.\n\n - `Face`: are sized 1/3 of the width of the bounding box;\n - `Torso`: upper biggest square croppable from the bounding box;\n - `Body`: square with side equal to the hight of the bounding box. If the limits of the image are exceeded (the image is too narrow), the first/last column of pixel is replicated to fill the gap;\n - `Bckg`: squares of 128 px are cropped randomly from the negative samples.\n\nAnd this is how they look like\n\n![Face](img/Face.jpg)\n![Torso](img/Torso.jpg)\n![Body](img/Body.jpg)\n![Bckg](img/Bckg.jpg)\n\n## How to run it\nAfter changing the folder pointers in the code (which I will make soon automatic), in the [`src`](src) directory, run\n\n```\nth makeClasses.lua\n```\n\n## How is this done?\nTo have an idea of the information provided with the dataset, I wrote [`checkBox`](src/checkBox.lua) which gives a general idea of what we can do with the data. To run it, go to the [`src`](src) directory and type\n\n```\nqlua checkBox.lua\n```\n\n![Bounding-box](img/Bounding-box.jpg)\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fatcold%2Ftorch-inria","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fatcold%2Ftorch-inria","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fatcold%2Ftorch-inria/lists"}