{"id":20058681,"url":"https://github.com/roboflow/magic-scissors","last_synced_at":"2025-06-24T23:07:54.138Z","repository":{"id":98131663,"uuid":"601390774","full_name":"roboflow/magic-scissors","owner":"roboflow","description":"Synthetic data for object detection and segmentation","archived":false,"fork":false,"pushed_at":"2023-10-05T20:17:02.000Z","size":898,"stargazers_count":12,"open_issues_count":4,"forks_count":3,"subscribers_count":6,"default_branch":"main","last_synced_at":"2025-06-19T03:11:28.743Z","etag":null,"topics":["augmentation","computer-vision","roboflow","synthetic-data"],"latest_commit_sha":null,"homepage":"https://magicscissors.app","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/roboflow.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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-02-14T00:47:02.000Z","updated_at":"2025-05-29T21:47:59.000Z","dependencies_parsed_at":null,"dependency_job_id":"4d209e26-0fdb-4704-9d7c-f47a896f4258","html_url":"https://github.com/roboflow/magic-scissors","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/roboflow/magic-scissors","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/roboflow%2Fmagic-scissors","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/roboflow%2Fmagic-scissors/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/roboflow%2Fmagic-scissors/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/roboflow%2Fmagic-scissors/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/roboflow","download_url":"https://codeload.github.com/roboflow/magic-scissors/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/roboflow%2Fmagic-scissors/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":261771127,"owners_count":23207220,"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":["augmentation","computer-vision","roboflow","synthetic-data"],"created_at":"2024-11-13T13:03:11.350Z","updated_at":"2025-06-24T23:07:54.113Z","avatar_url":"https://github.com/roboflow.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"![Magic Scissors](/assets/magic-scissors.jpg)\n\n# Magic Scissors\nGenerate synthetic data for computer vision using copy/paste context-augmentation.\n\nTry it: https://magicscissors.app\n\n## Background\nCollecting and annotating training to create performant computer\nvision models is hard time-consuming work. Often, the most valuable\ndata to collect is of the most rare edge-cases. It is impractical and,\nin some cases, impossible to wait for that data to present itself\nnaturally.\n\nCreating synthetic data to augment your real-world data can be\nextremely helpful in creating a robust dataset.\n\n## Context-Augmentation\nThere are several ways to expand your dataset (for example,\ntraditional data augmentation and creating 3D-rendered scenes).\nContext-augmentation is a simpler method which cuts objects of\ninterest from one scene and pastes them into another.\n\n## How to Use Magic Scissors\nMagic Scissors operates on [Roboflow](https://roboflow.com) datasets\nusing the [Roboflow API](https://docs.roboflow.com/rest-api). You'll\nneed [an API Key](https://docs.roboflow.com/rest-api#obtaining-your-api-key)\nand three datasets in your Roboflow account:\n\n1. `Objects of Interest` -- this should be an object detection or instance\nsegmentation dataset with the objects that should be cut out\n[annotated with polygons](https://docs.roboflow.com/annotate/the-labeling-interface#smart-polygons).\n2. `Backgrounds` -- this should be an object detection or instance\nsegmentation dataset with the backgrounds that objects will be pasted over.\nDrop-zones should be annotated with polygons.\n3. `Destination` -- this is where the new images will be stored. It should\nbe an object detection or instance segmentation project (and may, optionally,\nbe the same as your `Objects of Interest` dataset.\n\n## Settings\nThere are several settings available in Magic Scissors to customize your\noutput:\n\n- `Dataset Size` - the number of synthetic images to create.\n- `Objects Per Image` - how many objects to place on the background in\neach synthetic image (range of min to max).\n- `Object Size Variance` - the amount to grow or shrink objects of interest.\nThis simulates objects being closer or farther away from the camera and\ncan be used to adjust realism if the scale of the images in\n`Objects of Interest` and `Backgrounds` is not the same.\n\n## Applying Augmentations\nIt can be additionally useful to vary the color, brightness, contrast,\nrotation, etc of your objects of interest. You can choose these settings\non the `Objects of Interest` dataset in the Roboflow interface and Magic\nScissors will use them when isolating and exporting the objects.\n\n## Example Outputs\nExample projects will be listed below. If you create a public project\non [Roboflow Universe](https://universe.roboflow.com) with Magic Scissors,\nsubmit a PR to add it to this list.","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Froboflow%2Fmagic-scissors","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Froboflow%2Fmagic-scissors","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Froboflow%2Fmagic-scissors/lists"}