{"id":21151059,"url":"https://github.com/torinos-yt/nnonnx","last_synced_at":"2025-07-09T09:31:49.665Z","repository":{"id":63928346,"uuid":"571857332","full_name":"torinos-yt/NNOnnx","owner":"torinos-yt","description":"Using CUDA for Faster Machine Learning Inference on Unity","archived":false,"fork":false,"pushed_at":"2022-12-02T14:59:23.000Z","size":433,"stargazers_count":6,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"main","last_synced_at":"2024-11-15T00:32:52.208Z","etag":null,"topics":["cuda","machine-learning","onnxruntime","unity"],"latest_commit_sha":null,"homepage":"","language":"C#","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/torinos-yt.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":"2022-11-29T03:01:50.000Z","updated_at":"2023-06-16T03:05:21.000Z","dependencies_parsed_at":"2023-01-14T14:45:28.571Z","dependency_job_id":null,"html_url":"https://github.com/torinos-yt/NNOnnx","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/torinos-yt%2FNNOnnx","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/torinos-yt%2FNNOnnx/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/torinos-yt%2FNNOnnx/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/torinos-yt%2FNNOnnx/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/torinos-yt","download_url":"https://codeload.github.com/torinos-yt/NNOnnx/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":225531109,"owners_count":17483728,"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":["cuda","machine-learning","onnxruntime","unity"],"created_at":"2024-11-20T10:13:00.540Z","updated_at":"2024-11-20T10:13:01.291Z","avatar_url":"https://github.com/torinos-yt.png","language":"C#","funding_links":[],"categories":[],"sub_categories":[],"readme":"# NNOnnx\r\n\r\nThis is an alternative to [Barracuda](https://docs.unity3d.com/Packages/com.unity.barracuda@1.0/manual/index.html) for even faster machine learning inference on Unity in limited situations using the [onnxruntime](https://onnxruntime.ai/) and CUDA api.\r\n\r\nBy using [CUDA's Graphics Interoperability](https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__INTEROP.html) feature, NNOnnx uses resources on the GPU such as GraphicsBuffer and Texture directly as CUDA resources without copying them to the CPU. This is more useful for models that require higher resolution image input.\r\n\r\nThis provides the inference on more diversified onnx models and faster runtime speeds on PC platforms where CUDA available compared to the Unity Barracuda.\r\n\r\nNNOnnx is not intended to be a full wrapper around onnxruntime for Unity. If you want to use the full functionality of onnxruntime, use [Microsoft.ML.OnnxRuntime Nuget](https://www.nuget.org/packages/Microsoft.ML.OnnxRuntime/).\r\n\r\nSystem Requirements\r\n=================\r\n- Unity 2020.1 or higher\r\n- Windows: x64, D3D11\r\n- NVIDIA GPU\r\n- Path to DLLs contained in\r\n  - CUDA 11.x\r\n  - cuDNN\r\n  - TensorRT (Optionally, but most faster)\r\n\r\nFor more information on CUDA, cuDNN, and TensorRT version compatibility, please check [here](https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html) and [here](https://onnxruntime.ai/docs/execution-providers/TensorRT-ExecutionProvider.html). The easiest way to get these is to install the latest [Azure Kinect Body Tracking SDK](https://learn.microsoft.com/en-us/azure/kinect-dk/body-sdk-download).\r\n\r\nInstall\r\n=================\r\nNNOnnx uses the [Scoped registry](https://docs.unity3d.com/Manual/upm-scoped.html) feature of Package Manager for installation. Open the Package Manager page in the Project Settings window and\r\nadd the following entry to the Scoped Registries list:\r\n\r\n- Name: `torinos`\r\n- URL: `https://registry.npmjs.com`\r\n- Scope: `jp.torinos`\r\n\r\nNow you can install the package from `My Registries` page in the Package Manager\r\nwindow.\r\n\r\nRelated Project\r\n================\r\n- [Unity-TensorRT](https://github.com/aman-tiwari/Unity-TensorRT)\r\n- [TensorFlow Lite for Unity Samples](https://github.com/asus4/tf-lite-unity-sample)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftorinos-yt%2Fnnonnx","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftorinos-yt%2Fnnonnx","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftorinos-yt%2Fnnonnx/lists"}