- Видео 817
- Просмотров 1 344 325
Kevin Wood | Robotics & AI
США
Добавлен 23 ноя 2020
Subscribe to learn robotics, AI/ML, computer vision, and controls!
Robot Cat Toy: Pawsome Entertainment for Your Cat!
Worried your cat is too bored at home? Today we will be building a robot that keeps your cat entertained for hours! In this video, I’ll show you step-by-step how to create a simple robotic toy your cat will love. So grab your tools, and let’s get started!
Order Your BEAR Actuators: www.westwoodrobotics.io/bear-rfq/
Get Your 10% Limited Time Discount Code! Go to my website at www.kevinwoodrobotics.com, fill out your email when you see the pop up, and you will receive the discount code.
Thanks for watching! If you found this video helpful, please like, subscribe and share:
www.youtube.com/@kevinwoodrobotics
Sharing my referral link for when you order your Tesla. You’ll get $1,000 off the purch...
Order Your BEAR Actuators: www.westwoodrobotics.io/bear-rfq/
Get Your 10% Limited Time Discount Code! Go to my website at www.kevinwoodrobotics.com, fill out your email when you see the pop up, and you will receive the discount code.
Thanks for watching! If you found this video helpful, please like, subscribe and share:
www.youtube.com/@kevinwoodrobotics
Sharing my referral link for when you order your Tesla. You’ll get $1,000 off the purch...
Просмотров: 124
Видео
SAMURAI Robust Object Tracking without ANY TRAINING!
Просмотров 2,9 тыс.7 часов назад
Samurai just came out and it significantly improves on the visual object tracking performance from Meta’s SAM 2 model. Let’s see how robustly it tracks objects without any training! Previously SAM2 struggled with complex scenes and similar objects. SAMURAI overcomes these challenges with temporal aspects of its model and kalman filter to predict future object locations. See my review in this vi...
Automatic Number Plate Recognition (ANPR) using Python with YOLO and OCR
Просмотров 60216 часов назад
I will show you how to extract license plate number from an image using YOLO and OCR. Often times you'll here this as Automatic Number Plate Recognition (ANPR) or License Plate Recognition (LPR). 0:00 Introduction 0:13 Prepare License Plate Dataset 1:17 YOLO Training 2:36 YOLO Inference 3:10 OCR with PaddleOCR Thanks for watching! If you found this video helpful, please like, subscribe and shar...
PaddleOCR Python Tutorial: A Must-Try OCR Model for Image to Text!
Просмотров 45719 часов назад
PaddleOCR is a must-try ocr model to extract text from your images. Let’s go over how to use paddleocr in python! Thanks for watching! If you found this video helpful, please like, subscribe and share: www.youtube.com/@kevinwoodrobotics Sharing my referral link for when you order your Tesla. You’ll get $1,000 off the purchase of a Tesla product! ts.la/kevin145437 Social: Website: www.kevinwoodr...
Best Programming Languages for Robotics: Matlab vs Python vs C++!
Просмотров 631День назад
You could be wasting tons of time using the wrong programming language. Let's go over when to use matlab, python and C and the pros and cons of each one. 0:00 Introduction 0:17 MATLAB 1:37 Python 2:11 C 3:05 MATLAB vs Python vs C Summary Thanks for watching! If you found this video helpful, please like, subscribe and share: www.youtube.com/@kevinwoodrobotics Sharing my referral link for when yo...
Robotics Software Engineer Roadmap 2025! (Get Started in Robotics Today!)
Просмотров 2,1 тыс.14 дней назад
Are you trying to become a robotics software engineer? Whether you are transitioning into robotics from mechanical engineering, software engineering, a STEM major, or non-STEM major it is possible. With a decade of experience in robotics, I will walk you step-by-step through my 7 step process to take you from a beginner to an advanced robotics software engineer! Consulting (1-1): kevinwoodrobot...
Python Threading Tutorial: Basic to Advanced (Multithreading, Pool Executors, Daemon, Lock, Events)
Просмотров 2,7 тыс.21 день назад
Are you trying to make your code run faster? In this video, we will be taking a deep dive into python threads from basic to advanced concepts so that you can take advantage of parallelism and concurrency to speed up your program. 0:00 Introduction 0:53 What are threads? 1:18 When to use threads? (I/O vs CPU, GIL issue) 2:16 Python Thread without join() 3:10 Python Thread with join() 3:49 Python...
3D Object Detection (6D Pose Estimation) without Training using FreeZe
Просмотров 1,3 тыс.21 день назад
I will go over 3D object detection (or 6D pose estimation) without any training using Freeze from the paper Training-Free Zero-Shot 6D Pose Estimation with Geometric and Vision Foundation Models. Currently ranked 3 for 6D detection of unseen objects! 0:00 Introduction 0:14 What is 3D Object Detection (6D Pose Estimation)? 1:14 BOP: Benchmark for 6D Object Pose Estimation 1:49 Top 6D Object Dete...
Debug ROS 2 C++ Node with Breakpoint in VS Code by Running Node or Launch File (WSL and Ubuntu)
Просмотров 42221 день назад
I will show you how to debug your ROS 2 C node with breakpoints in VS Code for WSL and Ubuntu. 0:00 Introduction 0:14 Open VS Code (WSL Specific Method) 0:55 Install gdbserver and gdb 1:09 Colcon Build for Debug 1:28 Install VS Code C Extension 2:04 Create launch.json File 2:26 Debug C ROS Node with Breakpoint using Run 3:12 Debug C ROS Node with Breakpoint from Launch File Thanks for watching!...
Meta Robotic Hand: Feel with Vision Haptics! (Sparsh, Digit 360, Digit Plexus)
Просмотров 78721 день назад
Meta just came out with Sparsh, Digit 360, and Digit Plexus, which make up the robotic hand platform for tactile sensing with vision. Their Sparsh Model is capable of 6 main tasks (Force Estimation, Slip Detection, Pose Estimation, Grasp Stability, Textile Recognition, Bead Maze). The Digit 360 and Plexus make up the hardware and software to integrate the fingers and robotic hand. 0:00 Introduc...
Best Depth Estimation Models (MiDaS, Depth Pro, Depth Anything v2, DepthCrafter, Marigold, Metric3D)
Просмотров 2,3 тыс.28 дней назад
Which monocular depth estimation model is the best? I will review these common methods of depth estimation (Depth Anything V2, DepthCrafter, Marigold, Depth Pro, DPT/Midas, Metric3D) in terms of their specialty, speed, training availability and license. 0:00 Introduction 0:16 What is Depth Estimation? (Depth Map/Disparity vs Depth) 1:24 Depth Estimation Dataset 2:10 Depth Anything V2 2:48 Depth...
Best OCR Models to Extract Text from Images (EasyOCR, PyTesseract, Idefics2, Claude, GPT-4, Gemini)
Просмотров 3 тыс.Месяц назад
Which OCR (optical character recognition) model is best for text recognition? Let's compare the different models based on accuracy, speed, cost, local/non-local. I'll be comparing the different models like EasyOCR, PyTesseract, Idefics2, Claude, GPT-4, Gemini, Surya, DocTR. 0:00 Introduction 0:14 Why is OCR so challenging? 1:19 DEMO: OCR with LLM?! (PyTesseract vs EasyOCR vs GPT vs Claude) 2:34...
Meta Quantized Llama 3.2 1B and 3B! (FASTEST LLM Models in 2024?)
Просмотров 840Месяц назад
Meta AI just released the quantized 1B and 3B llama 3.2 model making it faster than ever and ready for mobile devices deployment for real-time usage. Thanks for watching! If you found this video helpful, please like, subscribe and share: www.youtube.com/@kevinwoodrobotics Sharing my referral link for when you order your Tesla. You’ll get $1,000 off the purchase of a Tesla product! ts.la/kevin14...
EasyOCR Python: Extract Text from Images with OCR (Improve Results with Image Processing)
Просмотров 711Месяц назад
I will show you how to use EasyOCR in python to do optical character recognition (OCR) to extract text from images and improve your results with some image processing techniques. Code and doc: kevinwoodrobotics.com/product/easyocr-python-extract-text-from-images-with-ocr-improve-results-with-image-processing/ 0:00 Introduction 0:17 What is OCR? 0:41 License Plate Dataset 1:19 Install EasyOCR 1:...
xAI API Tutorial: Run Grok Locally using Python on Windows
Просмотров 1,8 тыс.Месяц назад
xAI API Tutorial: Run Grok Locally using Python on Windows
Auto Label Images for YOLO Object Detection with AutoDistill (Annotate Images with Prompt)
Просмотров 1,4 тыс.Месяц назад
Auto Label Images for YOLO Object Detection with AutoDistill (Annotate Images with Prompt)
Top Challenges in Robotics (Haptics, Perception, AI, Design, Ethics)
Просмотров 152Месяц назад
Top Challenges in Robotics (Haptics, Perception, AI, Design, Ethics)
Medical SAM 2: Segment Medical Images as Video via Segment Anything Model 2
Просмотров 513Месяц назад
Medical SAM 2: Segment Medical Images as Video via Segment Anything Model 2
Labelimg: Image Labeling Tool for Object Detection (Annotate Images in MRI Tumor Dataset)
Просмотров 802Месяц назад
Labelimg: Image Labeling Tool for Object Detection (Annotate Images in MRI Tumor Dataset)
Football Player Tracking with SambaMOTR (Multi Object Tracking with Occlusion)
Просмотров 414Месяц назад
Football Player Tracking with SambaMOTR (Multi Object Tracking with Occlusion)
Football Player and Ball Detection with Custom Dataset using YOLO11 Object Detection
Просмотров 713Месяц назад
Football Player and Ball Detection with Custom Dataset using YOLO11 Object Detection
CoTracker3 Meta AI: ROBUST Tracking Under Occlusion and Scaling
Просмотров 1,1 тыс.Месяц назад
CoTracker3 Meta AI: ROBUST Tracking Under Occlusion and Scaling
Measure Size of Object in Images ACCURATELY using OpenCV Python
Просмотров 2,5 тыс.Месяц назад
Measure Size of Object in Images ACCURATELY using OpenCV Python
SAM2 Ultralytics (Setup and Code) Segmentation or Tracking
Просмотров 548Месяц назад
SAM2 Ultralytics (Setup and Code) Segmentation or Tracking
COCO Dataset Classes Index (Class ID) for Object Detection YOLO
Просмотров 151Месяц назад
COCO Dataset Classes Index (Class ID) for Object Detection YOLO
YOLO11 Ultralytics (Setup and Code) Object Detection Inference
Просмотров 600Месяц назад
YOLO11 Ultralytics (Setup and Code) Object Detection Inference
Will Robots Steal Our Jobs? (Thoughts After Seeing Tesla’s Optimus from We Robot)
Просмотров 444Месяц назад
Will Robots Steal Our Jobs? (Thoughts After Seeing Tesla’s Optimus from We Robot)
Depth Pro Apple: How Accurate is the Depth & Focal Length Estimation?
Просмотров 1,5 тыс.Месяц назад
Depth Pro Apple: How Accurate is the Depth & Focal Length Estimation?
We Robot Summary: Key Highlights in 5 Mins!
Просмотров 1,1 тыс.Месяц назад
We Robot Summary: Key Highlights in 5 Mins!
Tesla Robotaxi (Cybercab) Reveal from We Robot Event!
Просмотров 102Месяц назад
Tesla Robotaxi (Cybercab) Reveal from We Robot Event!
I've installed Everything and put them in the environment directories and tried 2 diff versions of Visual studio code and keep getting this error: [cmake] CMake Error at CMakeLists.txt:2 (project): [cmake] No CMAKE_CXX_COMPILER could be found.
Try to make sure you have a c++ compiler installed
How can I use Samurai? Is there a website or app?
Yeah they have a github
Hi. Can you help? I need to find the position and rotation of the TV screen in a photo, and I have a frame that is currently shown on this TV in good quality. But the problem is that the photo itself is of low resolution 320x240. Which algorithm is better to use from opencv?
You could find the lines on the tv and then calculate the angle
@kevinwoodrobotics thanks! I will try it.
Create a feedback loop with yolo to detect the bounding boxes of the cat and the toy to entertain the cats.
Oh like make it move away or towards the cat?
@kevinwoodrobotics now that I think, maybe you should store the feedback and profile the cat/cats : at which moment of the day they like to play, do they play both at the same time or separately, how do they "attack" the toy...
quick one, does this also do multi object tracking? I dont think that is the case but just confirming
Maybe you could by running multiple instances in parallel
Excuse me. I downloaded the file, unzipped it, and placed it in .nuke as you explained. But Cattery doesn't appear in Nuke. I don't know what's wrong! My nuke is 13.2v9.
I’m not following
CMake Error at CMakeLists.txt:2 (project): Running 'nmake' '-?' failed with: no such file or directory CMake Error: CMAKE_C_COMPILER not set, after EnableLanguage CMake Error: CMAKE_CXX_COMPILER not set, after EnableLanguage -- Configuring incomplete, errors occurred! I am facing this kind of problem. What can I do?
Make sure you have a c++ compiler installed
Next step: lift the cats up using 3D printed planetary gears.
😆
Greetings, excellent video, you can make a video of the installation of paddle ocr using the free models, I read the documentation it says that it works better with pyhon 3.12, the version of paddle 2.9.1
Thanks for the tip!
@@kevinwoodrobotics Thank you, I read the documentation, I don't know how PaddleOCR, PaddlePaddle and PaddleX are different, Although there is documentation on how to install it, I don't know how to give it pre-trained models that they say in the paddleOCR github since when I install it it doesn't appear. Greetings and success with your channel
Learn Robotics and AI: kevinwoodrobotics.com/product/all-courses-bundle-deal/ Order Your BEAR Actuators: www.westwoodrobotics.io/bear-rfq/ Get Your 10% Limited Time Discount Code for the BEAR Actuators (includes motor, encoder, motor driver)! Go to my website at www.kevinwoodrobotics.com/, fill out your email when you see the pop up, and you will receive the discount code.
Cat litter box next!!!
@@Lawring haha
Thank you so much!!!! it worked for me
Sweet!
will API's ever become available for models like these? I want to implement it in a drone or mobile robot to track people but have a GPU onboard isnt feasible. Thanks
They have a GitHub you can get started with
Can this handle multi-person pose estimation? if not which model would work best?
Technically yes, but have to test it out to see if performance is good enough for your application. Can also look into mocap methods
Really usefull for keeping up to date, new subscriber here :)
Awesome! Thanks for the sub!
ok but that player 26 movements tho
You are a amazing RUclipsr
Is this live performance
Not sure need to look into details
Rapid ocr built on top of paddle paddle ocr is a better option we use it for sub second with almost perfect accuracy every time
How do you build on top?
Does anyone here know how to get a conversation ID? Every time I talk to grok it creates a new conversation so it’s basically an amnesiac
Not sure but hopefully someone can answer this
Does it work for videos?
Yes. You can think of videos as many images (or frames), so you can analyze text frame by frame
Plz provide code any a live demo test ! For samurai .
They have some code on their github
Really like these new model updates! It's how I've been keeping up with things...thank you
Glad to hear!
Master Robotics and AI: kevinwoodrobotics.com/product-category/bundle-deals/
Dude. Nice content bro!
Thanks!
is this too hard to understand the math behind it?
May need some prerequisites like linear algebra
@@kevinwoodrobotics any suggestions on where to brush up those skills ?? Any video tutorial ?? As I am a visual learner ! 😢
Oh God it's that simple .....I took a lot of time to understand that's not that complicated , thank uu Sir
Glad to help!
what is your mail?, my method payment is paipal, send me message, i am interested your code
wwhere are k1 k2 and k3 of distortion matrix comes from?
Might have to look up the derivation for that
I have downloaded and unzipped, but now comes the tricky part for me. What is “pip install paddleocr paddlepaddle”? I'm on Windows 10.
This will be the command to run to install paddleocr
¡Qué buen video! Siempre me he preguntado cuál es el mejor lenguaje para la robótica. 🤖 Personalmente, he tenido experiencias interesantes con Python, pero me gustaría saber más sobre Matlab. ¿Alguien lo ha usado? Recientemente empecé a estudiar en сodigо hеroe y me está encantando 😊
can we plot the data stored in some file, like in .log file?
Can use Rosbags
@@shakeelshakeel8286 ROS2 Record Data And Playback Data ruclips.net/video/mFCundd5s-Q/видео.html
Thanks, Kevin. Another great video 🙂
Thank you! 😊
thanks for the video
Learn OpenCV: ruclips.net/p/PLSK7NtBWwmpQwSUi53XUK5o6-b9H3ABrO&si=plE7wzkDe8yxO9rP
wanna learn more!
Check out my playlist on my channel
ruclips.net/p/PLSK7NtBWwmpQwSUi53XUK5o6-b9H3ABrO&si=plE7wzkDe8yxO9rP
Hay quá ah ơi có vietsub nữa là oke
Thanks!
I will use Ubuntu to follow this tutorial, appreciate your effort <3
Awesome!
waiting for your video about vacuum gripper.
Computer Vision using AI Bundle DEAL: kevinwoodrobotics.com/product/computer-vision-using-ai/ Code: kevinwoodrobotics.com/product/automatic-number-plate-recognition-anpr-using-python-with-yolo-and-ocr/ Computer Vision Apps: kevinwoodrobotics.com/product-category/computer-vision-applications/
How do you use it fully locally like run the grok 1 model on your own device without using the api key to call xAI servers?
I don’t think that’s available at the time. Your best bet might be something like llama
Hello Kevin, Great Explaination, I tried the same concepts but for a Natural Image (Where i have an image as a marker and use the photo of marker clicked from my camera [Calibrated] and used as scene), I did the feature atching using SIFT and Flann and used SolvePnPRansac for find the good matches and then found the best rvectors and tvectors, i am projecting the points of the cuboid onto tho scene image (using ProjectPoints - OpenCV) but x and y are way too large, Any suggestions on how to soleve this issue? (All this is Without the ARUCO Marker)
Can you scale it down in the world coordinates before projecting?
@ Hey Kevin, how do we do that?
Good
is this Zhang method?
Yea
underrated, you gained a sub btw
Thanks!
Subbed. Looking fwd to using Hailo8 at the edge for LLM rather than vision.
Thanks!
I am really worried it makes mistakes.
What kind?
Great video! Super insightful and well-explained. 🙌 I'm currently looking for a computer vision engineer to join a project I'm working on. Do you have any advice on the best way to find the right talent for this? Would love to hear your thoughts!
Depends what you need specifically. But LinkedIn would be a good start
What is seed?
A point you choose
What if we click thw image form distance?? Will it give same resukts on it too? ? For that may be calibration needed cab you make a video of its whole process 😖😣
Yes that’s true
@kevinwoodrobotics please I need it badly. Kindly make a complete video for this. How can we meadure the size of object from different camera without doing calibration for every picture