Release AI-ModelZoo-3.0.0

The main features of the AI-ModelZoo-3.0.0 release:

• Model Zoo Services: Scripts to easily retrain, quantize, evaluate, or benchmark any model from user datasets, as well as application code examples automatically generated from user AI models, have been moved to the stm32ai-modelzoo-services GitHub.
• Additional Model Compatibility: Includes additional models compatible with the STM32N6570-DK board.
• Enhanced models offer: Expanded available models in all use cases.
• New Use Case Support: Expanded use case support to include Instance Segmentation and Speech Enhancement.
• PyTorch Support: Added PyTorch support through the Speech Enhancement use case.
• Model Zoo Hosting: Model Zoo hosted on Hugging Face.

Release AI-ModelZoo-2.1.0

The main features of AI-ModelZoo-2.1.0 release:

• Included additional models compatible with the STM32MP257F-DK2 board.
• Added support for per-tensor quantization.
• Integrated support for ONNX model quantization and evaluation.
• Included support for STEdgeAI (STM32Cube.AI v9.1.0 and subsequent versions).
• Expanded use case support to include Pose Estimation and Semantic Segmentation.
• Standardized logging information for a unified experience.

AI-ModelZoo-2.0.0

The main features of AI-ModelZoo-2.0.0 release:

• Aligned and uniformed architecture for all the use case folders.
• Modular design to run different operation modes (training, benchmarking, evaluation, deployment, quantization) independently or with an option of chaining multiple modes in a single launch.
• A single entry point to the code, with a single configuration file to configure all the modes.
• Added Bring Your Own Model (BYOM) feature.
• More training capabilities proposed (transfer learning, fine tuning, training from scratch).
• Addition of Object Detection models.

STM32AI ModelZoo 1.0.0

First release of STM32 Model Zoo includes 4 differents use-cases, a collection of pretrained models and Getting started packages for deployments.

Support for ST Developer Cloud Services.