Your object detector, benchmarked in DeepStream.
What it does
Imports an object-detection model into an NVIDIA DeepStream pipeline and produces inference engines, benchmarks, and a PDF report.
When to use it
Use when bringing a supported HuggingFace or NVIDIA NGC object-detection model into DeepStream; other model architectures are outside its scope.
How to use it
Choose the default detector or supply a custom model source; it then builds a DeepStream pipeline and writes benchmark reports into the working directory.
Uses
NN
Good to know
- acts without asking
Latest review
What you need · 4
Requires Docker with GPU access; Windows requires Docker Desktop with the WSL2 backend.
Requires the NVIDIA driver on the host.
Access · 4
This skill
HuggingFace
Read
Downloads models from Hugging Face
NVIDIA NGC
Read
Downloads models from NVIDIA NGC