Model Training

The final step of the workflow is computer vision model training and validation.

The NVIDIA TAO Toolkit is a low-code solution enabling developers and enterprises to accelerate model training and optimization processes. By using the annotated images created through synthetic data generation, you can fine-tune NVIDIA pre-trained models useful in computer vision use cases such as object detection, image classification, and segmentation. This reduces the barrier to entry for anyone starting with AI.

Model Training Examples

• Training Instance Segmentation Models Using Mask R-CNN on the NVIDIA TAO Toolkit

• Accelerating Model Development and AI Training with Synthetic Data, SKY ENGINE AI platform, and NVIDIA TAO Toolkit

• Bootstrapping Object Detection Model Training with 3D Synthetic Data

• Accelerating AI Development with NVIDIA TAO Toolkit and Weights & Biases

Deep Learning Institute

• Getting Started with Image Segmentation

• Getting Started with Deep Learning