Overview
The TI EdgeAI Model Hub is a curated repository of open-source computer vision models optimized for deployment on Texas Instruments MPU devices. Models are compiled for TI hardware using TIDL Runner or TIDL Tools, enabling production-ready inference without cloud dependency.
- Hardware-optimized for TI MPU processors
- Benchmarked on real TI hardware
- Format is portable and vendor-neutral
- Automated scripts for model preparation
Use Cases
Automotive · Aerospace & Defense · Industrial · Surveillance · Robotics · Edge IoT
License Summary
Models in this hub are distributed under various open-source licenses - each model's license is indicated in it's own documentation page.
Disclaimer: Certain licenses in this repository impose distribution restrictions that may affect commercial, proprietary, or regulated-industry use. It is the sole responsibility of the user to review the applicable license terms, assess compatibility with their intended use, and obtain any necessary legal clearances prior to use or distribution. Texas Instruments makes no representation regarding the suitability of these licenses for any particular purpose and accepts no legal responsibility for the user's compliance obligations.
Compilation & Deployment
| Tool | Description |
|---|---|
| TIDL Runner (recommended) | High-level compilation interface suitable for most users and standard deployment workflows. |
| TIDL Tools | Low-level compilation framework for advanced users and custom workflows. |
Quick Start
# 1. Clone the repository
git clone https://github.com/TexasInstruments/edgeai-modelhub.git
cd edgeai-modelhub
# 2. Navigate to a model directory and prepare the model
cd models/vision/<task>/<model>/
python prepare_model.py --model <variant>
# 3. Compile for TI hardware (run from inside the egdeai-tidlrunner directory)
cd /path/to/egdeai-tidlrunner
tidlrunner-cli compile --target_device <device> \
--config_path /path/to/egdeai-modelhub/<model>_config.yaml
# 4. Infer on TI hardware (run from inside the egdeai-tidlrunner directory)
cd /path/to/egdeai-tidlrunner
tidlrunner-cli infer --target_device <device> \
--config_path /path/to/egdeai-modelhub/<model>_config.yaml
Supported Hardware
Compatible TI MPU device families compiled and validated via TIDL. See the supported devices, SDKs and version compatibility at EdgeAI developer landing space and edgeai-tidl-tools sdk version compatibility matrix
| Device | Aliases |
|---|---|
| AM62A | — |
| J722S | TDA4AEN · AM67A |
| J721E | TDA4VM |
| J721S2 | TDA4VL · TDA4AL · AM68A |
| J784S4 | TDA4VH · TDA4AL · AM69A |
Model Catalog
| Model | Capability | Variants | Input | Performance | License | Repo |
|---|---|---|---|---|---|---|
| MobileNetV3 | Image Classification | large | 224×224 | 75.3% Top-1 | BSD-3-Clause | View |
| ResNet-50 | Image Classification | v1.5, v1 | 224×224 | 74.93–76.15% Top-1 | Apache 2.0 | View |
| DINO | Image Classification | ViT-S/16, ViT-S/8, ViT-B/16, ViT-B/8, ResNet-50 | 224×224 | 75.3–80.1% Top-1 | Apache 2.0 | View |
| DINOv2 | Image Classification | ViT-S/14, ViT-B/14 (w/ & w/o registers) | 224×224 | 80.9–84.6% Top-1 | Apache 2.0 | View |
| ViT | Image Classification | vit_b_16, vit_b_32, vit_l_16, vit_l_32 | 224×224 | 75.9–81.1% Top-1 | BSD-3-Clause | View |
| ConvNeXt | Image Classification | convnext_tiny, convnext_small, convnext_base, convnext_large | 224×224 | 82.5–84.4% Top-1 | BSD-3-Clause | View |
| DEIMv2 | Object Detection | s, m | 640×640 | 50.9–53.0% mAP | Apache 2.0 | View |
| DETR | Object Detection | detr_resnet50, detr_resnet50_dc5, detr_resnet101, detr_resnet101_dc5 | 800×800 (flexible) | AP50:95 42.0–44.9, AP50 62.4–64.7 | Apache 2.0 | View |
| Deformable-DETR | Object Detection | single-scale | 800×800 | AP50:95 39.4% | Apache 2.0 | View |
| RF-DETR | Object Detection | nano, s, m, l | 384–704px | 48.4–56.5% mAP | Apache 2.0 | View |
| RT-DETRv2 | Object Detection | s, ms, m, l, x | 640×640 | 48.1–54.3% mAP | Apache 2.0 | View |
| RTMDet | Object Detection | tiny, s, m, l, x | 640×640 | 40.9–52.8% mAP | Apache 2.0 | View |
| YOLO11 | Object Detection | n, s, m, l, x | 640×640 | 39.5–54.7% mAP | AGPL 3.0 | View |
| YOLO26 | Object Detection | n, s, m, l, x | 640×640 | 40.9–57.5% mAP | AGPL 3.0 | View |
| YOLOv8 | Object Detection | n, m | 640×640 | 37.3–50.2% mAP | AGPL 3.0 | View |
| YOLOX | Object Detection | nano, tiny, m, l, x, darknet53 | 416×416 / 640×640 | 24.8–51.2% mAP | Apache 2.0 | View |
Resources & Links
- Ecosystem: TI EdgeAI · EdgeAI SDK
- Tools: TIDL Runner · TIDL Tools
- Community: E2E Support · Issues · Discussions
Maintained by Texas Instruments EdgeAI Team | Last Updated August 2026