Added documentation for integration/deployment of models on TI Edge AI - #25403

All Contributors have signed the CLA. ✅

👋 Hello @nijil-ti, thank you for submitting a ultralytics/ultralytics 🚀 PR! To ensure a seamless integration of your work, please review the following checklist:

• ✅ Define a Purpose: Clearly explain the purpose of your fix or feature in your PR description, and link to any relevant issues. Ensure your commit messages are clear, concise, and adhere to the project's conventions.
• ✅ Synchronize with Source: Confirm your PR is synchronized with the ultralytics/ultralytics main branch. If it's behind, update it by clicking the 'Update branch' button or by running git pull and git merge main locally.
• ✅ Ensure CI Checks Pass: Verify all Ultralytics Continuous Integration (CI) checks are passing. If any checks fail, please address the issues.
• ✅ Update Documentation: Update the relevant documentation for any new or modified features.
• ✅ Add Tests: If applicable, include or update tests to cover your changes, and confirm that all tests are passing.
• ✅ Sign the CLA: Please ensure you have signed our Contributor License Agreement if this is your first Ultralytics PR by writing "I have read the CLA Document and I sign the CLA" in a new message.
• ✅ Minimize Changes: Limit your changes to the minimum necessary for your bug fix or feature addition. "It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is." — Bruce Lee

For more guidance, please refer to our Contributing Guide. This automated message confirms the initial review checklist; an Ultralytics engineer will assist with any follow-up questions. Don't hesitate to leave a comment if you need help. Thank you for contributing to Ultralytics! 🚀

I have read the CLA Document and I sign the CLA

Thanks for putting this together, and welcome — a TI Edge AI integration is genuinely interesting to us.

Direction first. The strongest version of this contribution is not a docs page. Every other silicon integration we ship — Rockchip RKNN, Sony IMX500, Axelera, DEEPX, Qualcomm QNN — is backed by a real export format in the package, so users get:

model.export(format="ti")

yolo export model=yolo26n.pt format=ti

As written, this page documents an entirely external toolchain: every step sends the user to edgeai-tidlrunner, prepare_model.py and a HuggingFace page, and nothing in ultralytics changes. That is a much weaker integration than TI deserves, and it is the kind of page that goes stale silently because no CI exercises it.

A ti export would slot into the existing machinery: add the row to export_formats() in ultralytics/engine/exporter.py, add an export_ti method (ONNX export + static-shape fixing + TIDL compile, which is what prepare_model.py already does), and wire a TIDL backend into ultralytics/nn/autobackend.py so YOLO("model_ti_model/") runs inference. Then the docs page documents something we own and test, and the workflow collapses to two commands. Happy to point you at the closest existing template to copy — RKNN and DEEPX are both good models for this.

If the docs page lands first, these need fixing:

1. The page is orphaned. It is not registered in mkdocs.yml nav, nor listed in docs/en/integrations/index.md, so nothing links to it and it will not appear in the site navigation.
2. The YOLO26 mAP row is wrong. The table lists 57.5 – 66.0; published YOLO26 detection is 40.9–57.5 mAP on COCO across n–x. 57.5 is the top of the real range, listed as the bottom. The YOLOv8 row also has "COCO trained" in the mAP column instead of a number.
3. Broken anchor. The FAQ links [Step 2](#step-2--compile-and-evaluate-with-tidl-runner), but that heading is "### Step 3 — Compile and Evaluate with TIDL Runner" — the anchor does not resolve.
4. "two-step pipeline" is followed by three numbered steps.

Still marked draft, so I have left it open. Ping me when you would like a review of either path.

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!