What is the output of bbox format in onnx file ? I want to use it on xavier with tensor rt export - Githubissues

lyuwenyu / RT-DETR

[CVPR 2024] Official RT-DETR (RTDETR paddle pytorch), Real-Time DEtection TRansformer, DETRs Beat YOLOs on Real-time Object Detection. 🔥 🔥 🔥

Apache License 2.0

2.61k stars 303 forks source link

What is the output of bbox format in onnx file ? I want to use it on xavier with tensor rt export #251

Open r4hul77 opened 7 months ago

r4hul77 commented 7 months ago

Star RTDETR 请先在RTDETR主页点击star以支持本项目 Star RTDETR to help more people discover this project.

Describe the bug A clear and concise description of what the bug is. If applicable, add screenshots to help explain your problem.

To Reproduce Steps to reproduce the behavior.

r4hul77 commented 7 months ago

Moreover, is there any flag to see output logs on tensorboard ?

lyuwenyu commented 7 months ago

[xmin, ymin, xmax, ymax]
Sorry, not yet. But I will add tensorboard in next version.

r4hul77 commented 7 months ago

太感谢了

r4hul77 commented 7 months ago

Btw, when I got the onnx file, it is expecting another input called "orig_target_sizes", what is this ?

gregspangenberg commented 7 months ago

Inferenced images will are padded up to the default 640x640 size. "orig_target_sizes" i believe is the sizes ('N', w , h) of the images you used as an input before it got padded by the model. The 'N' dimension must match the number of input images.

r4hul77 commented 7 months ago

ah that makes sense, I appreciate it.