{"id":30110321,"url":"https://github.com/leggedrobotics/qed-splatter","last_synced_at":"2025-08-10T04:44:28.120Z","repository":{"id":294273859,"uuid":"979535724","full_name":"leggedrobotics/qed-splatter","owner":"leggedrobotics","description":null,"archived":false,"fork":false,"pushed_at":"2025-08-06T10:11:01.000Z","size":114,"stargazers_count":1,"open_issues_count":1,"forks_count":1,"subscribers_count":9,"default_branch":"main","last_synced_at":"2025-08-06T12:10:02.835Z","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":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/leggedrobotics.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,"zenodo":null}},"created_at":"2025-05-07T17:01:43.000Z","updated_at":"2025-07-07T19:52:29.000Z","dependencies_parsed_at":"2025-05-19T17:54:03.616Z","dependency_job_id":"d4a6b4e1-51a9-4548-b059-c32adaa1146c","html_url":"https://github.com/leggedrobotics/qed-splatter","commit_stats":null,"previous_names":["leggedrobotics/qed-splatter"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/leggedrobotics/qed-splatter","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/leggedrobotics%2Fqed-splatter","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/leggedrobotics%2Fqed-splatter/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/leggedrobotics%2Fqed-splatter/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/leggedrobotics%2Fqed-splatter/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/leggedrobotics","download_url":"https://codeload.github.com/leggedrobotics/qed-splatter/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/leggedrobotics%2Fqed-splatter/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":269677516,"owners_count":24457858,"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","status":"online","status_checked_at":"2025-08-10T02:00:08.965Z","response_time":71,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":"2025-08-10T04:44:23.316Z","updated_at":"2025-08-10T04:44:28.103Z","avatar_url":"https://github.com/leggedrobotics.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# QED-Splatter\nQED-Splatter (short for Quick and Easy Depth Splatter) is a custom depth-supervised implementation of the Gaussian Splatting method, built on top of Nerfstudio. \nThe initial code is based on [DN-Splatter by maturk](https://github.com/maturk/dn-splatter) and hence parts of their code are used in this repository.\nCurrently [some branches](https://github.com/leggedrobotics/qed-splatter/tree/normal_testing) of this repo also work with the [DN-Splatter / AGS-Mesh](https://github.com/maturk/dn-splatter) Mesh-Extraction.\nIt is designed to provide a flexible and efficient framework for neural 3D scene reconstruction using depth information, particularly in outdoor and forested environments. \nThis repository was developed as part of a [Bachelor Thesis](https://github.com/leggedrobotics/forest-digital-twin) on applying Gaussian Splatting to forest environments.\n\n## Installation\nBefore installing QED-Splatter, make sure you have installed Nerfstudio following these [instructions](https://docs.nerf.studio/en/latest/quickstart/installation.html).\n```\nconda activate nerfstudio\ncd QED-Splatter\npip install -e .\nns-install-cli\n```\n\n## Setup\nPer default the `depth_unit_scale_factor` is set to millimeters. To update this change the `depth_unit_scale_factor: float` in the [dataparser](qed_splatter/dataparser.py).\n\nThe scaling to the 1 by 1 cube of nerfstudio is also enabled. For my thesis for example this needs to be disabled. To reenable this change add / comment in following lines in the [dataparser](qed_splatter/dataparser.py):\n```python\nauto_scale_poses: bool = False\ncenter_method: Literal[\"poses\", \"focus\", \"none\"] = \"none\"\norientation_method: Literal[\"pca\", \"up\", \"vertical\", \"none\"] = \"none\"\n```\n\nLastly when using unscaled scenes, the model currently doesn't initialize the gaussians based on the depth images. There are three options here.\n1. Use the default cube initialization and modify the scale of the cube with `random_scale: float = 100.0` in the [model config](qed_splatter/model.py).\n2. Use the 3d Points initialization provided by splatfacto.\n3. Add this functionality to the repository and submit a pull-request.\n\n## Running the new method\nTo train the new method, use the following command:\n```\nns-train qed-splatter --data [PATH]\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fleggedrobotics%2Fqed-splatter","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fleggedrobotics%2Fqed-splatter","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fleggedrobotics%2Fqed-splatter/lists"}