Update the bbox parser (yolov3)since the output format (model onnx) is different

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hello;

Please someone managed to implement the function which parses the output of model onnx to retrieve the bounding boxes (NvDsInferParseCustomYoloV3), I am using the library provided by deepstream5, but the bounding boxes are wrong
suggestions please
thank you

3

what do you mean by “wrong” ? number? location? Please elaborate.

4

location of bounding box

5

image
image847×377 143 KB

6

I suspect the output of your onnx model is not compabile with the YoloV3 parser in DeepStream. Did you check this?

Related Doc - Custom YOLO Model in the DeepStream YOLO App — DeepStream 6.1.1 Release documentation

7

yes already checked but even exits

8

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

can you share a repro?