Hello.
If I want for each bounding box to get all the probabilities for all the classes and not only for the “winner” class. Is there a way to do it somehow?
May I know which network did you mention?
Sure, I use Yolo v4 via nvidia tao
It is possible to get the confidence when run inference with the tensorrt engine.
For example, if you run inference via trition-app, you can take a look at the yolov3 inference via triton-app. From tao-toolkit-triton-apps/yolov3_postprocessor.py at main · NVIDIA-AI-IOT/tao-toolkit-triton-apps · GitHub , the confidence can be found for all the classes.
If you run inference with deepstream, you can refer to deepstream_tao_apps/nvdsinfer_custombboxparser_tao.cpp at master · NVIDIA-AI-IOT/deepstream_tao_apps · GitHub
How can i use Tao with yolor detection
Please create a new forum topic with more detailed info. Thanks.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.