![LearnOpenCV](/img/default-banner.jpg)
- Видео 329
- Просмотров 2 116 274
LearnOpenCV
США
Добавлен 13 фев 2015
Welcome to LearnOpenCV, a comprehensive RUclips channel dedicated to Computer Vision, Machine Learning, and Artificial Intelligence. Our mission is to provide high-quality educational content for everyone to succeed in these rapidly growing fields.
Our channel covers a wide range of topics, including deep learning, image processing, object detection, and face recognition, using state-of-the-art tools like OpenCV, PyTorch, and TensorFlow.
In addition to computer vision tutorials, we also offer valuable courses & career advice to help you achieve your professional goals
We encourage our viewers to engage with us by commenting, asking questions, and sharing their ideas. We want to create a collaborative learning environment where everyone can contribute and benefit.
Our channel is perfect for students, professionals, & hobbyists who are passionate about computer vision, ML & AI. Join us and start your journey towards becoming an expert in these exciting fields.
Our channel covers a wide range of topics, including deep learning, image processing, object detection, and face recognition, using state-of-the-art tools like OpenCV, PyTorch, and TensorFlow.
In addition to computer vision tutorials, we also offer valuable courses & career advice to help you achieve your professional goals
We encourage our viewers to engage with us by commenting, asking questions, and sharing their ideas. We want to create a collaborative learning environment where everyone can contribute and benefit.
Our channel is perfect for students, professionals, & hobbyists who are passionate about computer vision, ML & AI. Join us and start your journey towards becoming an expert in these exciting fields.
YOLOv10: A Leap Away from NMS - Advanced Object Detection Explained
📚 Blog post Link: learnopencv.com/yolov10/
📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/
In our latest video, we explore YOLOv10, a significant leap forward in the YOLO series, eliminating non-maximum suppression (NMS) and introducing several enhancements.
~ Overview of YOLOv10 and its six models, catering to different deployment needs from mobile to high-accuracy scenarios.
~ Detailed explanation of the new dual assignment method for NMS-free training.
~ Efficiency-accuracy driven model design, focusing on lightweight classification and critical regression heads.
~ Practical demonstration of running inference with YOLOv10 using a provided starter code....
📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/
In our latest video, we explore YOLOv10, a significant leap forward in the YOLO series, eliminating non-maximum suppression (NMS) and introducing several enhancements.
~ Overview of YOLOv10 and its six models, catering to different deployment needs from mobile to high-accuracy scenarios.
~ Detailed explanation of the new dual assignment method for NMS-free training.
~ Efficiency-accuracy driven model design, focusing on lightweight classification and critical regression heads.
~ Practical demonstration of running inference with YOLOv10 using a provided starter code....
Просмотров: 898
Видео
Instance Segmentation for Medical Imaging: YOLOv8 vs YOLOv9
Просмотров 50214 дней назад
📚 Blog post Link: learnopencv.com/yolov9-instance-segmentation-on-medical-dataset/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In our latest video, we finetune and compare YOLO instance segmentation models on a medical imaging dataset. ~ The significance of image segmentation, differentiating between semantic and instance segmentation with practical e...
Fine-Tuning YOLOv9: Experiment Results (Aerial Dataset)
Просмотров 44421 день назад
📚 Blog post Link: learnopencv.com/fine-tuning-yolov9/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In this video, we cover: ~ The experiments involved in fine-tuning a model and evaluating the fine-tuned YOLOv9 model's performance and inference results. ~ Using the SkyFusion: Aerial Object Detection dataset with 3 labels (aircraft, ship, vehicle) and s...
Stereo Vision in ADAS: Depth Perception Beyond LiDAR
Просмотров 54828 дней назад
📚 Blog post Link: learnopencv.com/adas-stereo-vision/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In this video, we cover: ~ The limitations of 3D LiDAR systems and how stereo vision can be used to extract 3D information for ADAS. ~ An introduction to stereo vision and training the stereo transformer model. ~ Capturing images from two slightly differe...
Integrating ADAS with Keypoint Feature Pyramid Network for 3D LiDAR Object Detection
Просмотров 415Месяц назад
📚 Blog post Link: learnopencv.com/3d-lidar-object-detection/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In this video, we cover: The significance of 3D LiDAR object detection in spatial understanding and its importance for domains like robotics and ADAS. The capabilities of 3D LiDAR data in estimating object volume, depth, shape, pose, trajectory pre...
LiDAR in ADAS: 3D Mapping and Environmental Perception
Просмотров 325Месяц назад
📚 Blog post Link: learnopencv.com/3d-lidar-visualization/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In this video, we explore the remarkable capabilities of LiDAR technology in Automatic Driver Assistance Systems (ADAS). We will guide you through a comprehensive code walkthrough using the 2D KITTI Depth Frames Dataset to create detailed 3D maps, enh...
Exploring Automatic Driver Assistance Systems (ADAS): Advancements and Impact
Просмотров 345Месяц назад
📚 Blog post Link: learnopencv.com/advanced-driver-assistance-systems/ 📚 Check out our FREE Courses at OpenCV University : opencv.org/university/free-courses/ In this video, we cover: The evolution of Automatic Driver Assistance Systems (ADAS), from Anti-lock Braking Systems to modern technologies like adaptive cruise control and drowsiness detection. The various levels of ADAS automation, from ...
Detect Home Intruders with OpenCV and Gradio
Просмотров 630Месяц назад
Detect Home Intruders with OpenCV and Gradio
Text Summarization using T5: Fine-Tuning and Building Gradio App
Просмотров 4072 месяца назад
Text Summarization using T5: Fine-Tuning and Building Gradio App
Everything You Should Know About a Camera: Internal Workings
Просмотров 3022 месяца назад
Everything You Should Know About a Camera: Internal Workings
Fine Tuning T5 for Stack Overflow Tag Generation
Просмотров 2342 месяца назад
Fine Tuning T5 for Stack Overflow Tag Generation
Deciphering LLMs: From Transformers to Quantization
Просмотров 2073 месяца назад
Deciphering LLMs: From Transformers to Quantization
Depth Anything: Accelerating Monocular Depth Perception
Просмотров 9863 месяца назад
Depth Anything: Accelerating Monocular Depth Perception
Mouse and Trackbar Functions in OpenCV | OpenCV Tutorial
Просмотров 3823 месяца назад
Mouse and Trackbar Functions in OpenCV | OpenCV Tutorial
A Guide to Annotating Images with OpenCV
Просмотров 7164 месяца назад
A Guide to Annotating Images with OpenCV
Reading and Writing Videos using OpenCV
Просмотров 4304 месяца назад
Reading and Writing Videos using OpenCV
Image Translation and Rotation Using OpenCV
Просмотров 3774 месяца назад
Image Translation and Rotation Using OpenCV
Image Filtering with Convolution in OpenCV
Просмотров 4674 месяца назад
Image Filtering with Convolution in OpenCV
OpenCV Simplified: Read, Display & Write Images | Beginner's Guide
Просмотров 7265 месяцев назад
OpenCV Simplified: Read, Display & Write Images | Beginner's Guide
Image Resizing with OpenCV - Tutorial
Просмотров 5515 месяцев назад
Image Resizing with OpenCV - Tutorial
Crop Images Like a Pro in OpenCV - Tutorial
Просмотров 6475 месяцев назад
Crop Images Like a Pro in OpenCV - Tutorial
Edge Detection Using OpenCV Explained.
Просмотров 1,3 тыс.6 месяцев назад
Edge Detection Using OpenCV Explained.
Contour Detection using OpenCV: A Comprehensive Guide
Просмотров 2 тыс.6 месяцев назад
Contour Detection using OpenCV: A Comprehensive Guide
Blob Detection in OpenCV & Python: A Comprehensive Guide
Просмотров 3,9 тыс.6 месяцев назад
Blob Detection in OpenCV & Python: A Comprehensive Guide
Fine-Tuning BERT using Hugging Face Transformers
Просмотров 9936 месяцев назад
Fine-Tuning BERT using Hugging Face Transformers
Fine-Tuning Segformer for Improved Lane Detection in Autonomous Vehicles
Просмотров 1 тыс.6 месяцев назад
Fine-Tuning Segformer for Improved Lane Detection in Autonomous Vehicles
Mastering BERT: An In-Depth Exploration of Revolutionary NLP
Просмотров 6627 месяцев назад
Mastering BERT: An In-Depth Exploration of Revolutionary NLP
Must one have installed fiftyone package?
No need to install it explicitly. The requirements.txt file will take care of all the necessary dependancies.
Tell me, can I use any lidar? Can livox Horizon be used for detection?
The underlying concepts are same. You are free to use any LiDAR.
@@LearnOpenCV Was there real-time work in the video? What about working in real time?
@@tomas111video The inference results shown as part of this work, is not real-time. The inference was performed on a pre-recorded video file. However, you can see that the model was about dishout ~180FPS on a RTX 3080 Ti FE GPU. We encourage you to try it on a real-time feed and report your results back.
@@LearnOpenCV Of course I'll try and let you know! Looking forward to new videos!
Thank you!
Finally we might get rid of NMS 😂
Wow❤
which is better?
You will find the answer to your question in the video,
Terrible, lazy video. Waste of time. This tells me any of the things any ML engineer should be interested in. Nothing about the model architecture, the training data, Hugging Face api, any technical detail, any performance benchmarks, nothing about re-training or transfer learing.. You just showed unreadable code snippets and waved your hands. What is the point? Useless.
Hey! I'm training in my computer and after one epoch I receive this message: The Kernel crashed while executing code in the current cell or a previous cell. Please review the code in the cell(s) to identify a possible cause of the failure. Click here for more info. View Jupyter log for further details.
does this works on languages other than english?
It work only in the English language.
Is there is any video while building it?
Can you provide the Code?
Please find the code in the download code section of the blog post!
Thank you! It's very simple explanation so I can understand it better
Parking space available app. Camera identifies open spots and provides "green shading" in real time to Parking app. Consider a large airport (tens of cameras providing GPS coordinates to GPS enabled phone). Same app could count cars providing subscription services on traffic volume (even parse by vehicle type by day to show more affluent customers to businesses catering to such clientele).
This could be a really good application!
Please, can you help me work lidar in cvat, is it possible?
Yes, it is possible. You will have to create a 3D task. Learn more here: www.cvat.ai/post/3d-point-cloud-annotation
@@LearnOpenCV Thanks 👍🏼🫂
in some of your jupyter notebook saw the code from clearml import Task from ultralytics import YOLO # Step 1: Creating a ClearML Task task = Task.init(project_name="my_project", task_name="my_yolov8_task") # Step 2: Selecting the YOLOv8 Model model_variant = "yolov8n" task.set_parameter("model_variant", model_variant) # Step 3: Loading the YOLOv8 Model model = YOLO(f"{model_variant}.pt"), tell me where this file I want to continue to perform steps on tutorial
Hi, which files are you referring to?
@@LearnOpenCV there was a phased implementation described in the article
@@LearnOpenCV there was a consistent description of what was described in the tutorial learnopencv.com/train-yolov8-on-custom-dataset/#results
Nice video. I would like to see more 3D content. Please, go on.
my google colab is not loading content in the file u provided
Nice.
it seems streaming on gradio isn't working
which is current SOTA Image Classification Technique?
Hey, you can check the latest SOTA papers and models on paperswithcode :)
Link in bio cannt be copied or clicked from the RUclips app, so we will never see the full video :/
Hi, here you go: ruclips.net/video/ZUhRZ9UTkIM/видео.html
Thank you for sharing this clear and simple information. Very easy to understand.
Thank you very much!
Tq
Great video sir!! Can you make a video demonstrating how to read text from a detecting box and then use the text in a database? Also the playlist on ADAS is really good, learning a lot about LiDARs
Great Video, thank you
Thank you for the appreciation.
@@LearnOpenCV You are Welcome 😊
Repository link
Please check the blog post link in the description.
I have a question. When aı try to export it, it is always in process. When is it gonna end?
If i click a create space it. Shows you have been rate-limted you can retry action later like that ....what can i do
Hi, did you try their suggestion? discuss.huggingface.co/t/rate-limit-when-using-gradio-and-inference-api/23508/2
On a side note. Can i use yolov8 for realtime custom object detection without using the ultralytics library?
No. YOLOv8 is part of the ultralytics library.
Hi sir, I have a segmentation annotation task, wherein I have 30 classes. The cvat application doesnt seem to be assigning a label value above 9. Is there some restriction in the number of classes we can annotate. How do i tackle this problem?
I think you should try deploying CVAT on your local system. You should not have problems then.
i run the code on 7 epochs and it always give me MaP : 0.00000e+00
when i use the code and train the model on 7 epochs , the MaP is always 0.00000e+00 , please give me a sulotion
Thank you so much for uploading this video on RUclips, if possible can you please make a video on how to use openPose along with open CV in a unity environment.
THANKS for this video, can you tell me how you get around the "catastrophic forgetting" notice when adding new classes and training data to an existing model?
So where are the OpenCV "tutorials" to get started on this topic?
We are in the process of making them. Please turn on your notifications to get intimation of the same.
Mediapipe does support multiperson detection now
Great to have an explanation at a much slower speed. Most machine learning tutorials go so fast I can't absorb the information. Also a much a more smoother transition between general description and advanced concepts. Not just throwing in advanced concepts suddenly, so the viewer has to stop playback and start looking things up elsewhere to keep up. I particularly like the way the narrator goes back and checks the viewer has picked up on a concept.
very nicely explained
Thank you!
I like this video! It answered a lot of questions I had as a beginner. Thank you so much! One question. This video is mainly about bounding box annotation. What about with key-point annotation? I am going to annotate mice in a cage, which means the objects are highly occluded. But I would like to use key-point annotation to detect their behaviour. What would be the best way to annotate to be consistent do you think?
We can use annotation tools such as imagelab, roboflow, etc for annotating keypoints
i thought it was in real time....
The app displays the output after every 24 frames. That's why it looks jittery.
Wow! Beginner want such class to understand. We want more of such tutorials.
Excellent tutorial! Thanks a lot...
Thank you so much!
how to train datasets using Yolov8 and SAHI?
Please refer to: github.com/obss/sahi
Thank you for posting this great tutorial. I was looking for it for quite a while. I am looking for a board to place inside my car. I want to use realtime object detection to recognise whenever a certain type of car is passing by in front of my car. And I want the board to notify me by 4g/lte sms( no wifi connection so i have to add a gsm/lte/4g module I am looking at 2 boards: 1. Raspberry pi 5 8Gb with Google Coral USB Accelerator 2. Google Coral Dev Board 4Gb Which one is the best option?
1 st one would be better option as it's more expandable.
Where are the visualized images stored? Where exactly in the memory? Where i can see them?
Dung iong i ham a lad, pait tang u tpu sain leh da ieng khiaw hi
how can ı use the tresholding on video
Hi, First create a VideoWriter object and read the video file, then extract the first video frame. Convert this frame to grayscale and then threshold it. You can watch videos on all these topics in this playlist: ruclips.net/p/PLfYPZalDvZDIm8OqtE7XiiLwq4ZyBVvhL
Is a GPU necessary for training the model?
Yes, GPU is necessary for training the model, however for inference depending upon the number of model parameters and size you can try with CPU accordingly.
For transfer learning, how many layers are there in total , so if the you froze the first 15, how many are being trained ?