Best Framework for Image Processing on Jetson Devices: Transitioning from DeepStream

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson
• DeepStream Version 6.2
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs) questions
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)


Currently, we have a functional pipeline running on DeepStream, specifically designed for image inference rather than video. However, we’ve encountered optimization challenges as DeepStream is more tailored for video streams. Our pipeline involves over 15 models, prompting the need for a more efficient approach.

To enhance our image processing capabilities on Jetson devices, we’re exploring alternatives to DeepStream. Considering our models are in both engine, ONNX, and TF formats, I’m seeking recommendations for a framework that is better suited for image processing on Jetsons.

Thank you!

I would recommend you using Titon Inference Server for this use case not deepstream sdk. The main advantage you can get from deep stream is stream processing acceleration and the tracking in videos. if you don’t process streams or videos then TRTIS is the best option. For building with Triton you can host multiple models and send http or gRPC requests directly to the server.