Skip to content
RobothecaBeta

Software

jetson-inference (Hello AI World)

From github.com/dusty-nv/jetson-inference (opens in a new tab)

  • Perception
  • Library
  • MIT

A C++ and Python library of deep learning vision models for NVIDIA Jetson devices, run with TensorRT and trained with PyTorch. It is also a tutorial called Hello AI World.

  • Closed core. It needs closed parts: NVIDIA TensorRT and JetPack.
  • Stale. No commit in 12 months: the newest is from 16 Oct 2025.
Open on GitHub (opens in a new tab)

Facts

GitHub stars
9,034read on 6 Nov 2026
Last update
16 Oct 2025Newest commit on master
Maintenance
Staleno commit in 12 months
Difficulty
Beginner
Hardware it needs, as the project states it
NVIDIA Jetson devices: Jetson Nano, Jetson Nano 2GB, Jetson Orin Nano, Jetson Xavier NX, Jetson AGX Xavier, Jetson AGX Orin, Jetson TX2 and Jetson TX1, each with JetPack and Ubuntu as listed in the system requirements. Training can also run on a PC, server or cloud instance with an NVIDIA discrete G

Checked on GitHub on 6 Nov 2026.

Licenses

Code
MIT (opens in a new tab)

Read by Robotheca from LICENSE.md (opens in a new tab) at commit 45da40a (opens in a new tab), on 16 Oct 2026. A license text Robotheca's reader does not know: a person at Robotheca set it on 9 Nov 2026.

Commercial use
Unknown

It needs closed parts under their own terms: read them before any commercial use.

One part it needs is closed

  • Needs NVIDIA TensorRT and JetPack: It runs its networks with NVIDIA TensorRT on Jetson boards, both under NVIDIA's own terms.

Overview

What it does: Robotheca's summary

  • Runs optimized vision networks on GPUs from C++ or Python using TensorRT.
  • Provides image classification, object detection, semantic segmentation, pose estimation, action recognition, background removal and monocular depth.
  • Includes live camera examples, WebRTC webapp examples and ROS/ROS2 support through a separate repository.
  • Lets you collect your own datasets and retrain classification and detection models with PyTorch.
  • Ships pre-trained models that are downloaded automatically when first used.
  • Its examples are compiled when the project is built from source.

Written by Robotheca from the project's README (opens in a new tab) at commit 45da40a, not by its creator.

The README

Read the README on GitHub (opens in a new tab)

Files and versions

Facts read at commit 45da40a

Robotheca never hosts the code: each link goes to the creator's own page.

Use it

How to install it, in the creator's words: The project's website (opens in a new tab)

Robotheca doesn't review code. Read it before running it.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.