search

DarshanAnand007 / YOLO_V8_Object_Detection

This is only for NVIDIA Jetson NANO
MIT License
4 stars 0 forks source link

You may do it faster 😉 #1

Open ViktorPavlovA opened 3 months ago

ViktorPavlovA commented 3 months ago

I have 8,9 fps on yolov8n with tensorRT+ cpp

DarshanAnand007 commented 3 months ago

what device are u running it

ViktorPavlovA commented 3 months ago

Jetson nano same

DarshanAnand007 commented 3 months ago

oh well i must be multi-threading, can you bried me about what exactly you are tring to run , can and what libraried and jetson have two model, i guess you can switch to max power for more fps

ViktorPavlovA commented 3 months ago

i don't test with threads, because i just start test yolov8 on jetson nano. In test i used model from COCO dataset. So test code is here https://github.com/ViktorPavlovA/tensorrt-yolov8-inference/tree/main. Model 640x640 , so output tensor [1,84,8400] , input [1,3,640,640]

DarshanAnand007 commented 2 months ago

ok good nice