a single dot or box

[xiangw369]

Thank you for providing, two onnx files. Would I like to ask, does this great algorithm only support the output of a single dot or box

[spacewalk01]

yes it supports a single point (dot) or a single box.

[xiangw369]

yes it supports a single point (dot) or a single box.

What should be done? Is it possible to output multiple boxes? Any ideas or suggestions?

[spacewalk01]

for multiple inputs, you can try this: https://github.com/NVIDIA-AI-IOT/nanosam