Hi, After converting a YOLO ( classifcation) model from ONNX to DeepStream format, I noticed that the model’s performance decreased significantly, from an F1 score of 98% to 67%! Why is this happening, and is there a method to correct the situation?

When converting to DeepStream, how do you run the evaluation, and are the preprocessing steps between DeepStream and when you run PyTorch the same?

Please provide complete information as applicable to your setup. • Hardware Platform (Jetson / GPU) • DeepStream Version • JetPack Version (valid for Jetson only) • TensorRT Version • NVIDIA GPU Driver Version (valid for GPU only) • Issue Type( questions, new requirements, bugs) • 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)

One reason this may occur is due to Ultralytics using a padding with (114,114,114), while DeepStream uses (0,0,0).

Thank you very much for your response @Y-T-G . I appreciate your insight on the padding issue. To resolve this problem and make the padding compatible between DeepStream and Ultralytics, do you suggest a specific approach or configuration to modify the padding values? Should I adjust the model preprocessing steps to ensure consistency, or would you recommend another solution? Looking forward to your guidance. Best regards

Thank you very much for your response @lqdisme. When I convert the model to DeepStream, I evaluate the performance by testing the model on the same test data (images) used to test the PyTorch model. I use the model generated with DeepStream and test this model with TensorRT.

I don’t think DeepStream supports padding with different color. The alternative is to modify Ultralytics to pad with 0,0,0

[image] amine.ghamgui.anavid: I use the model generated with DeepStream and test this model with TensorRT. do you mean the TRT engine generated with DeepStream is used in PyTorch test? please refer to this faq for “Debug Tips for DeepStream Accuracy Issue”.

Performance Drop After Converting YOLO ReLU6 Model from ONNX to DeepStream: Causes and Solutions

Accelerated Computing Intelligent Video Analytics DeepStream SDK

fanzh February 24, 2025, 8:32am 9

do you mean the TRT engine generated with DeepStream is used in PyTorch test?
please refer to this faq for “Debug Tips for DeepStream Accuracy Issue”.

Topic		Replies	Views
Performance Gap between Deepstream YOLO and pytorch YOLO DeepStream SDK	3	1120	January 2, 2024
Discrepancies in Object Detection Results between Ultralytics and DeepStream at Yolov8 DeepStream SDK	5	1246	February 26, 2024
Issue with Converting ONNX Model with different dimensions to TensorRT Engine for DeepStream DeepStream SDK deepstream	21	1026	May 23, 2025
Problems with Onnx model zoo -> trtexec -> DeepStream 6.0 pipeline DeepStream SDK	16	1211	November 29, 2023
No dection on video with yolo models on deepstream sdk DeepStream SDK yolo , jetson , deepstream	4	260	April 15, 2025
Can i use pytorch 1.6 with a trained yolo3 to use at deepstream5.1? Jetson Xavier NX	10	804	April 20, 2021
Classifier result on onnx doesn't match Deepstream result DeepStream SDK tensorrt , tensorflow , nvbugs , onnx	34	4099	July 28, 2021
Detections change in Deepstream 6.2 DeepStream SDK tensorrt , camera , cuda , kernel , gstreamer , yolo , python , deepstream , deepstream61	21	2266	July 31, 2023
YOLOv3 TensorRT model in DeepStream DeepStream SDK	8	1339	May 4, 2020
Loss of precision to onnx converter for engine by deepstream 6.3 DeepStream SDK tensorrt , gstreamer , inference-server-triton	30	3248	August 2, 2024

Performance Drop After Converting YOLO ReLU6 Model from ONNX to DeepStream: Causes and Solutions

Related topics