Model Composer is a fully integrated no-code solution for training and compiling models for deployment into edge AI applications. It supports Bring-Your-Own-Data (BYOD), enabling the re-training of models from the TI Model Zoo with custom data. Deploy to your local development board.

• Image classification
• Object detection
• Semantic segmentation
• Arc fault detection
• Motor bearing fault detection

For vision applications Model Analyzer allows you to experience first-hand the superior performance of TI’s optimized for edge AI products. It enables the evaluation of accelerated deep learning inference on remotely accessed development boards. Benchmark embedded deep learning inference in under five minutes. Select from hundreds of optimized, trained models or use a custom model. Obtain latency, frames-per-second processing, DDR bandwidth and accuracy benchmarks.

• AM62A starter kit (SK-AM62A-LP)
• AM68A starter kit (SK-AM68)
• AM69A starter kit (SK-AM69)
• TDA4VM starter kit (SK-TDA4VM)

For vision applications find the model that best meets your performance and accuracy goals on TI Processor from TI Model Zoo. Learn current performance statistics of models such as FPS, Latency, Accuracy & DDR bandwidth.

Model Maker is an advanced command-line end-to-end model development tool for vision applications. It is integrated into Model Composer, and is also available separately for command-line use on Linux® computers. It includes features for dataset handling, model training and compilation. Bring-Your-Own-Data (BYOD) is supported, which enables models from the TI Model Zoo to be re-trained with custom data.

Model Maker accepts annotated datasets from tools such as Model Composer and Label Studio.

• Image classification
• Object detection
• Semantic segmentation