YOLOv10 Comparison with Latest YOLO models
HTML-код
- Опубликовано: 11 сен 2024
- In this video, we test the speed 3 newest and best YOLO variants. To be more specific
YOLOv8-X
YOLOv9e
YOLOv10-X
Watch till the end to know the average FPS comparison.
🖥️ Computer Specifications
Processor : Intel Core i7
RAM : 16 GB
GPU : NVIDIA Geforce RTX 3060
GPU RAM : 12 GB
📱 Well, if you want to dive deeper YOLOv10, check out our 4-IN-ONE YOLO Course covering YOLOv10, YOLOv9, YOLOv8, and YOLOv7!
👉 www.udemy.com/...
#yolov10 #yolov9 #yolocomparison #yolov8 #objectdetection
#deeplearning #ai #yolo
Want to MASTER YOLOv10 Object Detection?
Check out our FOUR-IN-ONE YOLO Course covering YOLOv10, YOLOv9, YOLOv8, and YOLOv7!
👉 www.udemy.com/course/yolo-masterclass-deep-learning-computer-vision-course/?referralCode=0BD5279647173919231F
Good video. Yolov10 processes faster but is more imprecise.
Yolov8 and Yolov9 seem similar. I think Yolov8 is a little better with distant objects.
You have an accurate conclusion 👍
It seems like you are working with confidence above 0.2
I don't know if the footage is 720p or 1080p, can you please share the resolution?
I would like to test the performance of the nano versions of each model by using a 1080p and 2k images.
Object detection in this video uses a confidence threshold of 0.25.
The resolution is 1080p.
Hey, I am trying to make a conveyor belt tear detection model. Based on your comparison, which version would you recommend: YOLOv8, YOLOv9, or YOLOv10?
I have answered your question in another video
@@Dr.Priyanto.Hidayatullah YOLOv10 in your case. You would want the faster response.
@@Canna_Science_and_Technology yes
Nice, Can you please tell me how I will create the loss curve of yolov10, yolov9, and yolov8 for comparison on a single graph?.
That would probably a long tutorial 😬
V8 better pak, butuh yang lebih akurat di robot
Iya YOLOv8 lebih akurat. Namun di kasus tertentu, yolov9 lebih akurat.
Kalau ingin lebih akurat lagi, harus dilakukan modifikasi. Bisa arsitektur, hyperparameter, atau tambah data.
this is very slow for rtx3060
u should get around 700-800 fps if u quantized the model with int8 or int4
Thank you for the comment.
This is for fairness. I mean, I have to do quantization to all the model to be fair and it takes time.
At that time, I want to post the video asap and viewers sometime just want to know which one is better.
Btw, what quantization technic do you usually use?
@@Dr.Priyanto.Hidayatullah for quantization int4 should be fastest
i suggest u use a bit bigger model, so after quantz it will yield same performance
as none quantiz one
@@Dr.Priyanto.Hidayatullah yes thats good test.. but u need to explorer it more and push the efficenccy to max because u normally need to watch multiple streams, on the same time, which needs alot of computing power,
For that I agree. When deploying on a low spec hardware, we used any speed up technic possible to increase performance.
This video is a quick comparison between the three.