Yolov5 giving unexpected outputs

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) → aws T4
• DeepStream Version → 5.0
• TensorRT Version - 7.0
• NVIDIA GPU Driver Version (valid for GPU only) → 440.82

@ersheng Followed your yolov4 repo to make TRT engine for yolov5 which was built successfully. Compared the output with pytorch mode and they are both same. But when I hook it up in DeepStream I am not getting any boxes. I have uploaded the code and relevant files here . Let me know if you have any pointers!

Thanks

3

@y14uc339

Sorry for late reply.
How is status of your YOLOv5 deployment to DeepStream now?

4

@ersheng Haven’t been able to make progress, tried with pytorch → onnx → tensorrt → deepstream. But that too gives incorrect outputs. Precisely, too may bounding boxes just as in the case of when using tensorrtx/yolov5 to create deepstream app.

5

@y14uc339

I’ve given somebody who is also handling YOLOv5 some hints here. Hopefully they are also helpful for you.

7

Here’s a great tuto : GitHub - marcoslucianops/DeepStream-Yolo: NVIDIA DeepStream SDK 6.1 / 6.0.1 / 6.0 configuration for YOLO models

1 Like