Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

YouTube

NVIDIA DeepStream Technical Deep Dive - Multi-Object Tracker

Nvidia via YouTube

Overview

Dive deep into NVIDIA's DeepStream SDK and its multi-object tracking capabilities in this 28-minute technical video. Explore the fundamentals of the new tracker unified architecture, including how to select from three object tracker alternatives (NvDCF, DeepSORT, or IOU) or integrate your own tracker for vision AI app development. Gain insights into the tracker's state machine and learn which parameters can be configured to optimize performance for specific applications. Examine NVIDIA's state-of-the-art NvDCF tracker, comparing different configuration parameters and understanding the results and tradeoffs of various optimization strategies. Learn how to handle errors from detection and explore visual tracking in NvDCF. Conclude with a case study on TrafficCamNet, applying the concepts to real-world scenarios. Access additional resources for DeepStream tracker comparisons, parameter tuning, plugin documentation, and getting started guides to further enhance your understanding and implementation of multi-object tracking in AI-based multi-sensor processing applications.

Syllabus

Intro
DEEPSTREAM SDK
MULTI-OBJECT TRACKERS IN DEEPSTREAM
MULTI-OBJECT TRACKING IN A NUTSHELL
HANDLING ERRORS FROM DETECTION
VISUAL TRACKING IN NVDCF
CASE STUDY: TRAFFICCAMNET

Taught by

NVIDIA Developer

Reviews

3.0 rating, based on 1 Class Central review

Start your review of NVIDIA DeepStream Technical Deep Dive - Multi-Object Tracker

  • Profile image for Alex Barnadas
    Alex Barnadas
    Good video gathering, lack of exam or exercises or any original content. But the information in the videos is good, comes from Nvidia YT channel.

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.