Simple YOLOv8 Class for Object Detection with Webcam in Real-time

If you’re looking for courses and to extend your knowledge even more, check out this link here:
👉 www.nicos-school.com/ ➡️ get 20% OFF with the code: KZbin
In this video 📝 we are going to create a simple YOLOv8 class for Object Detection Model in Python. We will see how to deploy a trained YOLOv8 model and run live inference on a webcam. This can actually run in real-time. You can also check out my YOLOv7 Course where we cover everything from generating a dataset, setting up the model, training, and deployment.
If you enjoyed this video, be sure to press the 👍 button so that I know what content you guys like to see.
_______________________________________________________________
⚒️Freelance Work: www.nicolai-nielsen.com/nncode
_______________________________________________________________
🧑🏻‍💻 My AI and Computer Vision Courses⭐:
🤖 Enroll in CustomGPT Implementation Course: www.nicos-school.com/p/custom...
📚 Research Paper Implementation Course: www.nicos-school.com/p/resear...
📕 Object Detection Course:https:www.nicos-school.com/p/yolov7...
📗 OpenCV GPU Course:https:www.nicos-school.com/p/opencv...
📘 SegFormer Course: www.nicos-school.com/p/transf...
📙 Object Tracking Course: www.nicos-school.com/p/yolov8...
🦾 Online Courses with Job Guarantee on Springboard (Save $1000 with: "NICOLAINIELSEN") www.springboard.com/landing/i...
_______________________________________________________________
📞 Connect with Me:
🌍 My Website: www.nicolai-nielsen.com/
🤖 GitHub: github.com/niconielsen32
👉 LinkedIn: / nicolai-h…
🐦 Twitter: / nielsencv_ai
_______________________________________________________________
⌛TimeStamps⏳
Timestamps:
0:00 Intro
0:42 YOLOv8 Class
4:00 Load Model
5:53 Predict
6:29 Extract Info & Visualize
11:56 Main Loop
14:37 Results
Tags:
#yolov8 #yolo #objectdetection #ultralytics

[@NicolaiAI]

📘Enroll in Research Paper Implementation Course: nicolai-nielsen-s-school.teachable.com/p/research-paper-implementation 📗Enroll in Object Tracking Course:nicolai-nielsen-s-school.teachable.com/p/yolov8-object-tracking-course 📕Enroll in YOLOv7 Course:nicolai-nielsen-s-school.teachable.com/p/yolov7-custom-object-detection-with-deployment

[@totuan4276]

Thank you, it’s so clear and detail.

[@NicolaiAI]

Glad you like it! Thanks for watching

[@maiquelkappel7745]

Congratulations on the video. Could you make a video like this, but using monitor screen?

[@seyedamirhh]

Thank you so much, this is really helpful and well explained.

[@NicolaiAI]

Glad it was helpful! Thanks for watching

[@al-musbahitechnology5766]

Thank you, you helped me a lot, but I still have a problem. I trained a model to recognize license plates. It identified them, but I also want it to print the number of the car that it identified. Can you explain this method?

[@cashyc6314]

Can this implement in raspberry pi? Will this makes the fps drops?

[@a.a.ardebili7433]

Thank you for the video. How I can set a filter to detect and draw BBox ONLY for "Person"? Is it possible to set up an area in the video frame the detector detects the objects only in that area?

[@NicolaiAI]

Thanks a lot for watching! Yeah u can just crop out your region of interest. And if u check for index 0 in the classes it corresponds to a person. I have another video here on the channel where I implement Yolov8 in a class for person detection

[@user-qm8ix3cn7k]

I want to transfer the names of the detection objects in yolov8's real time analysis to the mobile application in real time via websocket, how do I do that?

[@inteli3i5i7]

The following error comes up: ValueError: too many values to unpack (expected 4) It is in the following lines: self.labels = [f"{self.CLASS_NAMES_DICT[class_id]} {confidence:0.2f}" for _, confidence, class_id, tracker_id in detections]

[@user-sp6iy5kn4e]

hi did it get solved or not. Because it also gives me same value error . please help if you solve this error

[@ombothre2350]

Same error

[@AbdelrahmanOsama74]

Try to make this change , it solved the problem for me Just add another '_' after 'for' self.labels = [f"{self.CLASS_NAMES_DICT[class_id]} {confidence:0.2f}" for _,_, confidence, class_id, tracker_id in detections]

[@umarmohamed2183]

Thanks for that Video, it's really helpful. if possible, can you share the name of your chair plz😅.

[@NicolaiAI]

Thanks a lot for watching! My chair in the background is from IKEA and my Office chair is a Herman Miller embody

[@vijaypaulraj9177]

How can we print the objecct detected?

[@athanasioschatzipanagiotou8119]

Great video. However I wonder how to get a very smoothly output. I have tried multithreading however the output is a bit of choppy. Not tottaly smooth. Do you have any solution on that

[@NicolaiAI]

Thanks! Are u thinking about tracking? I have a course about that with yolov8

[@mystuff6071]

Great Video! But I unfortunately had an error occurred, "TypeError: unsupported format string passed to list.__format__". Any idea on how to fix this?

[@NicolaiAI]

In what line do u get the error?

[@mystuff6071]

@@NicolaiAI Nevermind, got it fixed! Keep up the great work though! Your course on YOLO v8 must be really in depth if you can afford to teach us this for free!

[@NicolaiAI]

@@mystuff6071 there is also some theory in the course and how to do everything from scratch. Also different ways to deploy the model

[@mystuff6071]

Hi, I would like to know how to count the amount of detected object per class. Is this possible? I could only get the amount of all the detected objects.

[@NicolaiAI]

U could have a list with all the counters and then increment the index in that list based on the label’s index

[@user-nn7pn9cp8d]

where is the code can you share me the git link of your repo pls

[@fassesweden]

If I only want to display bboxes with confidence >0.5 where Can I do this in the code? Great video!!

[@NicolaiAI]

U Can basically just have an if statment checking the confidence score and then only visualize or store if greater than threshold

[@albertmatevosyan303]

@@NicolaiAI Hey, I couldn't implement said statement into your code. Is there any way you can help?

[@neiltonna8737]

@@NicolaiAI In which method should this if statement be made please? Cannot understand a way where to put this statement.

[@thekushalgaikwad]

Can you build an app which integrates yolov8 for object detection?

[@NicolaiAI]

Will def do that in the future

[@nhatpham5797]

Hello, thank for your video But i have a problem: from supervision.tools.detections import Detections, BoxAnnotator ModuleNotFoundError: No module named 'supervision.tools' Can you fix it? Thank you, have a nice day!

[@NicolaiAI]

Hi Thanks for watching! Did u pip install supervision?

[@nhatpham5797]

@@NicolaiAI Yupp, i installed it, But it still error

[@nhatpham5797]

@@NicolaiAI In my supervision library there is no "tool" module but there is still "draw" module, I don't understand why

[@adrianknutsen]

same problem

[@nhatpham5797]

@@adrianknutsen i fixed it, please remove "supervision" version 0.2.0 and install "supervision" version 0.10

[@JasonKim-yz8rv]

Can YOLOv8 perform instance segmentation in real-time?

[@NicolaiAI]

Yeah for sure! I actually have a video about that aswell. Even for a custom dataset

[@JasonKim-yz8rv]

@@NicolaiAI Wow... I found it. That's brilliant!!!

[@JasonKim-yz8rv]

kzbin.info/www/bejne/apq1d3qgg9mthNE

[@NicolaiAI]

@@JasonKim-yz8rv great to hear! Hope u can use it

[@user-so9en8wz5b]

I came here after seeing your comment link. I'm trying to load my trained best.onnx file. However, I'm getting the following error: return self._sess.run(output_names, input_feed, run_options) onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Got invalid dimensions for input: images for the following indices index: 2 Got: 640 Expected: 320 index: 3 Got: 640 Expected: 320 Please fix either the inputs or the model. I tried adding self.img_size = (320, 320) in __init__ and modifying the predict method as follows: def predict(self, frame): frame = cv2.resize(frame, self.img_size) results = self.model(frame) return results But I still encounter the same error. Is there any specific reason for this?

[@NicolaiAI]

U are trying to use onnx with PyTorch. U can’t really do that since onnx is a whole another framework. In my yolov7 course I go over how to use custom trained yolo models with onnx framework

[@user-so9en8wz5b]

@@NicolaiAI kzbin.info/www/bejne/i3-5eo14gJlooc0 Is the video in question correct? Thank you for your prompt response!

[@NicolaiAI]

Yeah that’s the course I’m talking about. Then u can do inference with ONNX

[@currogonzalezfleta7151]

Hi, I have the next error: ColorPalette.__init__() missing 1 required positional argument: 'colors'. Any idea?

[@Nourhan-jw8iv]

Did u know how to fix this error

[@neiltonna8737]

Has this error. Use 'sv.ColorPalette.default()' instead.

[@RAGAVNGSEC]

it says there is no modele named supervision help!

[@NicolaiAI]

Did u pip install it?

[@christoragav860]

@@NicolaiAI thankss

[@yaseenpk1739]

After pip installing. It says there us no module named supervision.tools

[@TechAirSpace]

@@yaseenpk1739 Did you figure this out? I'm getting the same error.

[@filipbarbaric1960]

why do you have two cameras.

[@NicolaiAI]

Some people might find it interesting to see how it is behind the scenes

[@Erolsaurus]

Hi, I got ValueError: too many values to unpack (expected 4) on line 68

[@artemtelushkin186]

Hi, did you solve this problem?

[@ombothre2350]

Same error

[@AbdelrahmanOsama74]

Try to make this change , it solved the problem for me Just add another '_' after 'for' self.labels = [f"{self.CLASS_NAMES_DICT[class_id]} {confidence:0.2f}" for _,_, confidence, class_id, tracker_id in detections]