YOLOX is a high-performance anchor-free YOLO, exceeding yolov3~v5 with MegEngine, ONNX, TensorRT, ncnn, and OpenVINO supported. Documentation: https://yolox.readthedocs.io/
亲爱的大老您好
因为我的边缘装置不支援 exp
我希望改成sigmoid 或 pow(2)
所以我把 decode 和 output的地方更换
但是收敛效果都不太好
无法达到 exp 的效果
是否能指点我
谢谢您
Hello dear boss
Because my edge device doesn't support exp
I hope to change it to sigmoid or pow(2)
So I replaced the decode and output places
But the convergence effect is not very good
Unable to achieve the effect of exp
Can you give me some advice?
thank you
亲爱的大老您好 因为我的边缘装置不支援 exp 我希望改成sigmoid 或 pow(2) 所以我把 decode 和 output的地方更换 但是收敛效果都不太好 无法达到 exp 的效果 是否能指点我 谢谢您 Hello dear boss Because my edge device doesn't support exp I hope to change it to sigmoid or pow(2) So I replaced the decode and output places But the convergence effect is not very good Unable to achieve the effect of exp Can you give me some advice? thank you