{"name": "SamGIS", "package_name": "samgis", "description": "\u26a0\ufe0f Archived / unmaintained. This project is part of the samgis-* line, which I'm no longer actively maintaining. Maintaining it has been unpaid work, and without real user feedback to justify the continued time investment, I'm redirecting my effort toward projects with a sustainable return. The code stays available as-is (MIT) \u2014 issues and PRs may go unreviewed. Forks are welcome, see https://gitlab.com/aletrn/samgis-qgis-plugin/\r\n\r\n-----------------\r\n\r\nLocal SAM2 instance segmentation for QGIS. Draw points and rectangles on a vector layer, run Segment Anything Model 2 inference on a raster layer, and get polygon results \u2014 all offline with ONNX Runtime. Supports multiple SAM2.1 model variants (Tiny to Large, fp32 and uint8). More features coming soon (live canvas preview, multi-band support, and more).", "about": "SamGIS brings Meta's Segment Anything Model 2 (SAM2) to QGIS as a local inference plugin. No cloud API needed \u2014 models run on-device via ONNX Runtime. Use foreground/background point prompts and bounding-box prompts drawn on vector layers to segment objects from any georeferenced raster. Results are added as polygon vector layers in EPSG:4326. Model weights are downloaded on first use from Hugging Face (aletrn/sam2-onnx-weights).", "homepage": "https://gitlab.com/aletrn/samgis-qgis-plugin", "repository": "https://gitlab.com/aletrn/samgis-qgis-plugin", "tracker": "https://gitlab.com/aletrn/samgis-qgis-plugin", "author": "trincadev", "tags": ["raster", "machine learning", "segment", "segmentation", "ai", "onnx", "sam", "sam2", "object detection", "interactive"], "downloads": 1453, "latest_version": "1.2.1", "versions": [{"version": "1.2.1", "experimental": false, "qgis_min": "3.40.0", "qgis_max": "4.99.0", "downloads": 1300, "uploaded_by": "trincadev", "upload_datetime": "2026-03-17T12:20:41.812927"}, {"version": "1.2.0", "experimental": false, "qgis_min": "3.40.0", "qgis_max": "4.99.0", "downloads": 68, "uploaded_by": "trincadev", "upload_datetime": "2026-03-17T10:14:20.282608"}, {"version": "1.1.1", "experimental": false, "qgis_min": "3.40.0", "qgis_max": "4.99.0", "downloads": 85, "uploaded_by": "trincadev", "upload_datetime": "2026-03-14T06:33:01.464605"}]}