.svg)
.svg)
Access a large library of production-tested NVIDIA® TAO models and let the Edge Impulse EON (Edge Optimized Neural) Tuner find the perfect fit for your application.
The Edge Impulse platform will enable you to select the edge-optimized NVIDIA TAO model for your use case and deploy on any efficient, low-cost edge devices, including MCU, CPU, and accelerators.
Compare model architectures with different size and latency with the EON Tuner and choose
the right one for any edge application.
Choose the most performant NVIDIA TAO model with on-device performance profiling and be sure it will run efficiently on the most optimized production-ready hardware.
.webp)
Leverage the EON Tuner and the ability to profile on device performance on any edge hardware to confidently make the right trade-off between model size, accuracy and latency to achieve your business goals.
With the Edge Impulse and NVIDIA TAO Toolkit, professional engineers can create accurate, custom, production-ready computer vision models that can be seamlessly deployed to SWaP (Size, Weight and Power) optimized edge AI hardware.
With the Edge Impulse and NVIDIA TAO Toolkit, professional engineers can create accurate, custom, production-ready computer vision models that can be seamlessly deployed to SWaP (Size, Weight and Power) optimized edge AI hardware.
With the Edge Impulse and NVIDIA TAO Toolkit, professional engineers can create accurate, custom, production-ready computer vision models that can be seamlessly deployed to SWaP (Size, Weight and Power) optimized edge AI hardware.
With the Edge Impulse and NVIDIA TAO Toolkit, professional engineers can create accurate, custom, production-ready computer vision models that can be seamlessly deployed to SWaP (Size, Weight and Power) optimized edge AI hardware.
We've selected models that will fit on a wide range of hardware, including:
Edge Impulse platform can output libraries in the following formats: C++11, TFLite, WebAssembly, TensorRT, or specialized binaries for AI accelerators. In addition, the Impulse can be exported as a Linux process inside a Docker container or as a separate process that can be controlled via Node.js, C++, Go, or Python.
In the first release we are adding support for both object detection and image classification using NVIDIA TAO models. For object detection, we support 80+ unique TAO model possibilities.
Every embedded vision task is unique, so we support a range of different model architectures. Here are the heads and backbone options that can be fine-tuned to your dataset and hardware constraints.
.svg)
Quickly solve industrial business problems with the hardware you know and the tools you need.
Contact Sales
Synthetic data and testing environments enable robust models with 10x faster results.
Low volume edge AI development is now possible on NVIDIA production ready hardware.
.svg)
