Welcome to the future of computer vision with Savant, an open-source, high-level framework that empowers developers to create efficient multimedia AI applications. In this blog, we will guide you through how to set it up and get you started on your journey towards building real-time, highly efficient applications on the Nvidia stack.
Why Choose Savant?
Savant revolutionizes the development of computer vision solutions by offering dynamic, fault-tolerant inference pipelines that harness the power of Nvidia technologies. This allows you to build advanced video analytics applications, all while maintaining an easy-to-use and scalable framework.
Getting Started with Savant
Follow these steps to get your Savant development environment up and running:
- Clone the Savant repository from GitHub:
git clone https://github.com/insight-platform/Savant.git - Navigate to the Savant directory:
cd Savant/samples/peoplenet_detector - Use the following command depending on your hardware:
- For x86 systems:
utils/check-environment-compatible && docker compose -f docker-compose.x86.yml up - For Jetson systems:
utils/check-environment-compatible && docker compose -f docker-compose.l4t.yml up
- For x86 systems:
- Open your RTSP stream:
- Via your player:
rtsp://127.0.0.1:554/stream/city-traffic - Or visit: http://127.0.0.1:888/stream/city-traffic
- Via your player:
- Press Ctrl+C to stop the running docker-compose bundle.
- Return to the project root:
cd ..
Understanding Savant with an Analogy
Imagine you are a chef preparing a complex dish. You have numerous ingredients, cooking tools, and techniques at your disposal. Savant is your culinary assistant; it organizes your kitchen (the framework) and preps each ingredient (the data inputs) in a manner that you can easily whip up your delicacies (multimedia AI applications) without worrying about the technical intricacies of the cooking process.
Just as this assistant enables you to focus on creating a delightful meal, Savant allows developers to focus on building intelligent applications without delving into the underlying complexities of the technology stack.
Troubleshooting Savant Issues
While using Savant, you may encounter certain issues or challenges; here are a few troubleshooting tips:
- Ensure that you are using the recommended versions of Nvidia DeepStream and JetPack hardware. Check the compatibility tables provided in the README.
- If your pipelines are not exhibiting the expected performance, verify the configuration in the runtime guide. Adjust any settings accordingly.
- Restart your Docker containers if you see any discrepancies in the output.
For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.
Final Thoughts
At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.
Join Us on This Journey
Savant is designed for those who are keen on implementing high-performance, production-ready computer vision applications. Whether you’re working on edge devices or across a data center, Savant provides a solid foundation that is flexible, scalable, and efficient.
Next Steps
Feel free to dive deeper by exploring:
Happy coding!