What's Changed

• Add SAHI logo to README and docs by @fcakyon in #1366
• docs: ✨ add custom styles for light and dark themes in extra.css by @onuralpszr in #1367
• 👷 Bump actions/checkout from 6.0.2 to 6.0.3 by @dependabot[bot] in #1370
• Fix numba parity mismatches with numpy and torchvision by @ZephyrKeXiner in #1372
• 👷 Bump astral-sh/setup-uv from 8.1.0 to 8.2.0 by @dependabot[bot] in #1369
• package: 📦 version bump sahi to 0.12.1 by @onuralpszr in #1373

Full Changelog: 0.12.0...0.12.1

Sliced Aided Hyper Inference detect the small stuff in big images.

This is one of the largest SAHI releases to date: 95 commits since 0.11.34 (rolling in the 0.11.35/0.11.36 hotfixes), spanning a re-architected post-processing engine, true batch inference, a lighter torch-free core, six new model families, and a full documentation overhaul.

🚀 Highlights

⚡ Batch inference, torch-free core & accelerated post-processing backends (#1336)

A ground-up reworking of how SAHI runs and merges predictions:

• Batch inference slices are now processed in batches end-to-end, dramatically improving throughput on GPU.
• Torch-free core the core slicing/ post-processing path no longer hard-depends on PyTorch. Install only what your model backend needs.
• Pluggable post-processing backends NMS / NMM now run on a selectable backend:
  • NumPy zero heavy deps, runs anywhere
  • Numba JIT-accelerated CPU path for large slice counts
  • TorchVision GPU-accelerated when torch is available
  • Backend auto-selects based on what's installed, and can be forced explicitly.

🎚️ Finer control over slicing & post-processing

• force_postprocess_type in get_sliced_prediction for explicit control over how overlapping detections are merged (#1346).
• Per-call confidence_threshold override across the prediction APIs tune confidence without rebuilding the model object (#1352).
• Progress bar + progress callback for get_sliced_prediction, available in both the Python API and CLI (#1255).

🧠 New model support

• GroundingDINO (HuggingFace) zero-shot, text-prompted open-vocabulary detection. Describe what you're looking for in natural language and run it through SAHI's sliced pipeline perfect for finding small, rare, or unlabeled objects in large images ships with a dedicated demo notebook (#1361).
• Universal segmentation from HuggingFace run any HF universal/panoptic segmentation model through SAHI's sliced pipeline (#1360).
• RF-DETR-Seg segmentation variant of RF-DETR (#1315).
• YOLOE detection model (#1268).
• YOLO-World open-vocabulary detection model (#1267).
• YOLO26 support across the Ultralytics backend, CLI, docs, and notebooks (#1321, #1322, #1356).

📚 Documentation overhaul

• Migrated docs to Zensical with full code typing & formatting cleanup (#1344).
• Chinese (zh) translation of the documentation added and kept in sync (#1253, #1332, #1347).
• New API reference, ** post-processing backends guide**, security policy, and Code of Conduct (#1257, #1336, #1272, #1349).

✨ Performance & Improvements

• Significantly faster post-processing NMS, NMM, and GREEDYNMM now use a shapely STRtree spatial index, dramatically speeding up merging on images with many slices/detections (#1248). Thanks @nikvo1!
• Faster read_image_as_pil for quicker slicing throughput (#1353).
• Improved performance & resource management in prediction and slicing (#1263).
• Better nms performance with correct handling of empty predictions (#1288).
• Replaced pybboxes with a lightweight in-house yolo_bbox_to_voc_bbox (#1320).
• Dropped the pybboxes and pinned opencv-python version constraints for cleaner installs (#1325).

🐞 Bug Fixes

• Fixed empty bounding boxes caused by an empty shapely_annotation.multipolygon (#1140).
• Fixed invalid segmentation masks for Detectron2 models (#1262).
• Corrected margin calculation in BoundingBox (#1286).
• Fixed CHW-format image handling in read_image_as_pil (#1287).
• Validate overlap ratios in get_slice_bboxes (must be < 1.0) (#1285).
• Corrected error message for invalid model path in RTDetrDetectionModel (#1266).
• Fixed incorrect type annotations in the postprocess module (#1327).
• Ultralytics model now supports additional formats with improved task handling (#1321).
• Added pywinpty for Windows dev compatibility (#1319).

🧹 Maintenance & CI

• Pinned all GitHub Actions to commit SHAs for supply-chain security (#1351).
• Multi-OS CI matrix and clearer workflow naming (#1334).
• Bumped to Python 3.12/3.13 in CI and docs (#1259, #1260).
• Removed deprecated YOLOv5 helpers, legacy requirements.txt, MMDet workflow, and unused Netlify config (#1326, #1342, #1341, #1335).
• numpy<3.0, torchvision 0.23.0, and many dependency bumps via Dependabot (now also covering pip).

🙏 Contributors

Thanks to everyone who contributed to this release:
@onuralpszr, @fcakyon, @siromermer, @ZephyrKeXiner, @yogendrasinghx, @srikrishnavignesh, @ibuldakov, @ducviet00, @volks73, @RizwanMunawar, @nikvo1, @vinnik-dmitry07, and @gboeer and @dependabot for keeping dependencies fresh.

🌟 New Contributors

• @vinnik-dmitry07 made their first contribution in #1140
• @nikvo1 made their first contribution in #1248
• @volks73 made their first contribution in #1262
• @ZephyrKeXiner made their first contribution in #1253
• @yogendrasinghx made their first contribution in #1326
• @ibuldakov made their first contribution in #1315
• @srikrishnavignesh made their first contribution in #1360
• @siromermer made their first contribution in #1361

Full Changelog: 0.11.34...0.12.0

What's Changed

This release torch import error in cli

• release: ✨ sahi version to 0.11.35 in pyproject.toml (#1252) by @onuralpszr in #1252

Full Changelog: 0.11.35...0.11.36

What's Changed

• Fix currently failing CI by adding a hatch build target by @gboeer in #1245
• refactor: ♻️ improve readability and type hints across codebase by @onuralpszr in #1249
• Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #1244
• refactor: ♻️ update CI workflow for multi-OS support and improve job naming by @onuralpszr in #1250
• Fix empty bbox because of empty shapely_annotation.multipolygon by @vinnik-dmitry07 in #1140
• Significantly improve postprocessing speed (NMS, NMM, GREEDYNMM) by using STRtree from shapely by @nikvo1 in #1248
• fix: 🐞 add optional dependencies and improve package management in models and late import for torch to fix ci by @onuralpszr in #1251

New Contributors

• @vinnik-dmitry07 made their first contribution in #1140
• @nikvo1 made their first contribution in #1248

Full Changelog: 0.11.34...0.11.35

What's Changed

• docs: add mkdocs material badge for support by @onuralpszr in #1236
• Fix: Models always loaded on "cuda:0" when working inside subprocesses on multi-GPU setups by @malopez00 in #1230
• refactor: Update category remapping logic to work with immutable categories by @gboeer in #1234
• feat: add support for torchvision Mask R-CNN instance segmentation model by @curtiskennedy in #1238
• refactor: make exif rotation on image read by default and use PIL built-in ImageOps instead by @fcakyon & @onuralpszr in #1239
• build: Bump version from 0.11.33 to 0.11.34 by @fcakyon in #1241

New Contributors

• @malopez00 made their first contribution in #1230
• @curtiskennedy made their first contribution in #1238

Full Changelog: 0.11.33...0.11.34

What's Changed

• docs: Simplify and improve context7 config by @fcakyon in #1218
• build: Eliminate unnecessary CI workflow triggers by @fcakyon in #1219
• fix: Fix incorrect context7 config structure by @fcakyon in #1220
• refactor: 🧹 improve test and utility code organization by @onuralpszr in #1223
• 👷 Bump actions/checkout from 4 to 5 by @dependabot[bot] in #1229
• fix: 🐛 Update version to 0.11.33 to fix sahi installation in dev by @onuralpszr in #1233

Full Changelog: 0.11.32...0.11.33

What's Changed

• docs: 📝 Introduce mkdocs-material to sahi project to have documentation page by @onuralpszr in #1210
• 👷 Bump astral-sh/setup-uv from 5 to 6 by @dependabot[bot] in #1211
• fix: 🐞 update deployment branch from 'develop' to 'main' in publish_docs workflow by @onuralpszr in #1212
• docs: 📝 Introduce mkdocs-material to sahi project to have documentation page by @onuralpszr in #1213
• Add DeepWiki documentation badge in README.md by @RizwanMunawar in #1215
• refactor: 🛠 remove redundant logging imports and centralize logger configuration by @onuralpszr in #1214
• update version to 0.11.32 by @fcakyon in #1216

New Contributors

• @onuralpszr made their first contribution in #1210
• @dependabot[bot] made their first contribution in #1211
• @RizwanMunawar made their first contribution in #1215

Full Changelog: 0.11.31...0.11.32

What's Changed

• Make Category immutable and add tests by @gboeer in #1206
• Update docstring for greedy_nmm by @kikefdezl in #1205
• update version by @fcakyon in #1208

New Contributors

• @kikefdezl made their first contribution in #1205

Full Changelog: 0.11.30...0.11.31

SAHI v0.11.30 Release Notes

We're excited to announce SAHI v0.11.30 with improved performance tracking, enhanced testing
infrastructure, and better developer experience!

📈 Milestones

• Academic papers citing SAHI reached 400! (#1168)

🚀 Key Updates

Performance & Monitoring

• Fixed postprocess duration tracking in get_sliced_prediction - now properly separates slice,
  prediction, and postprocess timings for accurate performance monitoring
  (#1201) - Thanks @Toprak2!

Framework Updates

• Refactored Ultralytics support with ONNX model support and better compatibility
  (#1184)
• Updated TorchVision support to latest API (#1182)
• Improved Detectron2 support with better config handling to prevent KeyError issues
  (#1116) - Thanks @Arnesh1411!
• Added Roboflow framework support for RF-DETR models from the Roboflow Universe
  (#1161) - Thanks @nok!
• Removed deepsparse integration as the framework is no longer maintained
  (#1164)

Testing Infrastructure

• Migrated test suite to pytest (#1187)
  • Tests now run faster with better parallel execution
  • Extended Python version coverage (3.8, 3.9, 3.10, 3.11, 3.12)
  • Updated to more recent PyTorch versions for better compatibility testing
  • Improved test organization and maintainability
• Refactored MMDetection tests for better reliability (#1185)

Developer Experience

• Added Context7 MCP integration for AI-assisted development
  (#1198)
  • SAHI's documentation is now indexed in Context7 MCP
  • Provides AI coding assistants with up-to-date, version-specific code examples
  • Includes llms.txt file for AI-readable documentation
  • Check out the Context7 MCP installation
    guide to integrate SAHI docs with your AI
    workflow

🛠️ Improvements

Code Quality & Safety

• Immutable bounding boxes for thread-safe operations (#1194,
  #1191) - Thanks @gboeer!
• Enhanced type hints and docstrings throughout the codebase
  (#1195) - Thanks @gboeer!
• Overloaded operators for prediction scores enabling intuitive score comparisons
  (#1190) - Thanks @gboeer!
• PyTorch is now a soft dependency improving flexibility
  (#1162) - Thanks @ducviet00!

Infrastructure & Stability

• Improved dependency management and documentation (#1183)
• Enhanced pyproject.toml configuration for better package management
  (#1181)
• Optimized CI/CD workflows for MMDetection tests (#1186)

🐛 Bug Fixes

• Fixed CUDA device selection to support devices other than cuda:0
  (#1158) - Thanks @0xf21!
• Corrected parameter naming from 'confidence' to 'threshold' for consistency
  (#1180) - Thanks @nok!
• Fixed regex string formatting in device selection function
  (#1165)
• Resolved torch import errors when PyTorch is not installed
  (#1172) - Thanks @ducviet00!
• Fixed model instantiation issues with AutoDetectionModel.from_pretrained
  (#1158)

📦 Dependencies

• Updated OpenCV packages from 4.10.0.84 to 4.11.0.86 (#1171) -
  Thanks @ducviet00-h2!
• Removed unmaintained matplotlib-stubs dependency (#1169)
• Cleaned up unused configuration files (#1199)

📚 Documentation

• Added context7.json for better AI tool integration (#1200)
• Updated README with new contributors (#1175,
  #1179)
• Added Roboflow+SAHI Colab tutorial link (#1177)

Acknowledgments

Special thanks to all contributors who made this release possible: @nok, @gboeer, @Toprak2, @Arnesh1411,
@0xf21, @ducviet00, @ducviet00-h2, @p-constant, and @fcakyon!

Full Changelog: 0.11.24...0.11.30

What's Changed

• Make bounding box immutable by @gboeer in #1194
• Improve type hints and docstrings by @gboeer in #1195
• update version by @fcakyon in #1196

Full Changelog: 0.11.28...0.11.29