My pipeline runs smooth but the bounding boxes that should be formed around the detected objects but they just get collected at the corner of the window for no reason.
Hello @sidd.aiindia,
Are you following a specific set of instructions to setup your Deepstream inference demo ?
Are you familiar with the DeepStream-Yolo repo? If not, you might find it very useful.
best regards,
Andrew
Embedded Software Engineer at ProventusNova
Please refer to deepstream_tools/yolo_deepstream/deepstream_yolo at main · NVIDIA-AI-IOT/deepstream_tools
Thanks proventusnova for replying on my query.
I have searched the DeepStream-Yolo repo that you shared with me. I created the pipeline to run the application. But, I got the same output.
Thank you for your help.
Siddhesh
Embedded AI Developer at AI India Innovations
Do you mean you got the bboxes in the corner of the image with the samples in deepstream_tools/yolo_deepstream/deepstream_yolo at main · NVIDIA-AI-IOT/deepstream_tools · GitHub? With the model we provided or with your own model?
Thanks Fiona.Chen for replying on my query.
I referred your provided repo, the pipeline I build by following the commands and process I got a fully working pipeline solving the error I got of misplaced bounding boxes with your model.
But the detection accuracy is low, so can you help me increase the accuracy?
I didn’t mean yours.
What kind of accuracy issue?
For the accuracy of the model, it is a training issue.
If the model’s accuracy is good but runs poor with DeepStream, it may be a configuration or algorithm issue.
There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks.