Train Yolov8 object detection custom data in the cloud GPU | AWS project | Computer vision tutorial

Code: github.com/computervisioneng/...
🎬 Timestamps ⏱️
0:00 Intro
0:26 Project overview
1:25 Data
2:44 Create S3 bucket
4:13 Create SNS topic
6:16 Create EC2 instance
10:21 Install drivers and requirements
15:31 Code
38:35 Results
39:00 Outro
🌍 Community 👥
Join our Discord server: / discord
Support me on Patreon: / computervisionengineer
#python #yolov8 #objectdetection #aws #awsprojects #computervision

[@ComputerVisionEngineer]

Try my premium courses! 😃🙌😊
● End-To-End Computer Vision: Build and Deploy a Video Summarization API bit.ly/3tyQX0M
● Hands-On Computer Vision in the Cloud: Building an AWS-based Real Time Number Plate Recognition System bit.ly/3RXrE1Y
● Machine Learning Entrepreneur: How to start your entrepreneurial journey as a freelancer and content creator bit.ly/4bFLeaC
All my premium courses are available to the Computer Vision Experts in my Patreon. 😉
www.patreon.com/ComputerVisionEngineer

[@andrewkarp]

@ComputerVisionEngineer thank you for this detailed video, much appreciated!

[@albertwang5974]

Thanks for sharing such kind of useful process!

[@ComputerVisionEngineer]

You are welcome! Glad it was helpful! 😃🙌

[@mayankmalhotra3523]

Hey Felipe! I've been really enjoying your computer vision videos on CZcams, they're super insightful! 🚀 I was wondering if you could point me in the right direction for videos on building detection models without using any type of pretrained models. Any recommendations or insights would be greatly appreciated! Keep up the fantastic work! 👍

[@ComputerVisionEngineer]

Thank you! Glad you find the content helpful! Sure, there are many ways in which you can train an object detector without using any type of pre trained models. With yolov8, for example, you can train from an existing pre trained model or you can also train a model from scratch. 🙌

[@guillermovelazquez3211]

Hola Felipe, has explorado las herramientas que ofrece Ultralytics para controlar todo el ciclo de vida del modelo?, por ejemplo como iniciar con pocas imágenes y utilizar el modelo para hacer predicciones y seguir refinando el modelo, al mismo tiempo usar algun logger como wandb, cometML, etc. para monitorear las versiones y métricas del modelo, así como también el deployment donde podemos usar el formato ONNX (que Ultralytics dice que es más rápido al ejecutarse en CPU), TensorRT (para optimizar en GPU nvidia), o algún otro formato de exportación. También cosas como el método tune que te permite encontrar los mejores hyperparámetros para tu modelo, o herramientas que se acoplan bien con yolov8 como la librería supervision que tiene muchas utilidades al momento de hacer el plot de los objetos detectados o tracking, yo he explorado por encima estas cosas que te menciono pero me falta un poco organizar tantas ideas y posibilidades, quería saber si tu has tenido oportunidad de probarlas debido a que has sacado muy buenos videos respecto a YOLOv8. Un saludo, y muchas gracias por tus videos que son de tanta utilidad, que estes bien Felipe!!

[@ComputerVisionEngineer]

Hola, no he tenido la oportunidad de explorar esas herramientas, pero trataré de hacerlo pronto! 😃🙌

[@user-hc9lp5wr5e]

hi i want to you make the tutorial how we can use the yolov8 model to make inference using gpu on the aws sagemaker

[@ComputerVisionEngineer]

Ok, I will try to. 🙌

[@user-firebender]

is there a way to detect object from video and return/output text that have the timestamp and the object label?

[@ComputerVisionEngineer]

Yes, it is possible to make a script to detect objects from videos and return the timestamp the object was detected. 🙌

[@user-firebender]

@@ComputerVisionEngineer i have been searching it for a few days but haven't found anything other than yolov5 version and that one that says to edit the model.py file, do you have any clue/tutorial to print the timestamp of object in txt?

[@pratham1972]

Hi Felipe,I want to make weapon Detection Computer Vision Project .Which can detect Knifes,Pistol,any other weapons in real time.Can you Tell me how to do this .If possible share me any authenticate resources you have if any .
Thanks

[@ComputerVisionEngineer]

Hi, once you have the data you can follow the process I describe in this video. Try looking for publicly available datasets. 🙌

[@sanjoetv5748]

Hello.. can you teach us how to train yolov8 on a nvidia graphics card

[@ComputerVisionEngineer]

This video is about training yolov8 on a GPU in a EC2 instance in AWS. If not mistaken the ec2 instance we used in this tutorial (p2xlarge) uses an NVIDIA K80.

[@SujithReddy-hb1wy]

hey how can we deploy on aws sagemaker gpu ?

[@ComputerVisionEngineer]

Hi, do you mean how to deploy yolov8 in sagemaker gpu?

[@sujithreddy1599]

Yess

[@ComputerVisionEngineer]

@@sujithreddy1599 I will look into it 🙌

[@andrewkarp]

Anyone else encountering this error "Unable to locate package Nvidia-driver-450

[@andrewkarp]

Solved it, I created a server with an incorrect Ubuntu version.