Hello, i have trained a yolov7-segmentation and converted it to TensorRT, but i recently found that there ir another branch called mask which actually works for instance segmentation as well. I would like to know if there is a difference in accuracy/speed since it seems that i would need more frames.
Anyone has experience whit this?
If not, i would upload some benchmarks using Nvidia edge devices in these days, so stay tuned xd
Another question, do you know if there is a tiny version for the segmentation task? i've tried to modify the configuration file but it seems that it is not as easy as that, im getting even worse results.
Hello, i have trained a yolov7-segmentation and converted it to TensorRT, but i recently found that there ir another branch called mask which actually works for instance segmentation as well. I would like to know if there is a difference in accuracy/speed since it seems that i would need more frames. Anyone has experience whit this? If not, i would upload some benchmarks using Nvidia edge devices in these days, so stay tuned xd
Another question, do you know if there is a tiny version for the segmentation task? i've tried to modify the configuration file but it seems that it is not as easy as that, im getting even worse results.
Thanks.