Videos

Our YouTube channel, Edje Electronics, focuses on democratizing access to computer vision technology by providing easy-to-follow projects and tutorials on using TensorFlow, OpenCV, and more.

 

MaskCam: How to Set Up your NVIDIA Jetson Nano as a Web-Connected Mask Detection Smart Camera
24:46
Edje Electronics

MaskCam: How to Set Up your NVIDIA Jetson Nano as a Web-Connected Mask Detection Smart Camera

Want to turn your NVIDIA Jetson Nano into a web-connected smart camera? This video shows how to set up MaskCam, a mask detection camera implemented with NVIDIA Deepstream, YOLOv4-tiny, and TensorRT, on your own Jetson Nano. It also shows how to set up a local Docker server on your computer that can be used to remotely interact with the Nano and view detection statistics. I worked on this project with BDTI, a technology analysis firm in the Bay Area. The project can serve as a reference design for other connected smart camera applications. It also gives a good example of how to use Docker containers for setting up web databases and using MQTT to communicate detection data. --- Links --- MaskCam Part 1 video, which gives an overview of the hardware and software behind this camera: https://youtu.be/W7SxGhJp8CA GitHub repository: https://github.com/bdtinc/maskcam Link to NVIDIA GTC 2021 conference talk about MaskCam (you need to register to see the talk, but it's free and available on-demand): https://gtc21.event.nvidia.com/media/t/1_etsb5xdn/204677953 MaskCam informational report by BDTI: https://bdti.com/maskcam Link on how to use barrel jack power supply input: https://www.jetsonhacks.com/2019/04/10/jetson-nano-use-more-power/ Amazon Associate link to Jetson Nano: https://amzn.to/3gRTkDc Amazon Associate link to 5V, 4A power supply: https://amzn.to/3t6KP9C --- Music --- Morning Station by Tokyo Music Walker https://soundcloud.com/user-356546060​ Creative Commons — Attribution 3.0 Unported — CC BY 3.0
How to Use the Coral USB Accelerator with the Raspberry Pi - Increase TensorFlow Lite FPS!
08:37
Edje Electronics

How to Use the Coral USB Accelerator with the Raspberry Pi - Increase TensorFlow Lite FPS!

Want to achieve blazing fast detection speeds (30+ FPS) with your TensorFlow Lite models on the Raspberry Pi? This video shows how to set up Google's Coral USB Accelerator with the Raspberry Pi to run TFLite object detection models on videos or real-time webcam feeds. It works for the Pi 3B+ and the Pi 4. Have questions? Ask me on Twitter @EdjeElectronics ! I usually respond faster there: https://twitter.com/EdjeElectronics -- Affiliate Links -- Get a Rasbperry Pi 4: https://amzn.to/2Kf0el8 Coral USB Accelerator from Google: https://amzn.to/2wxTZ8d Coral USB Accelerator from seeed studio (sometimes less expensive): https://amzn.to/3bBVY9Z Webcam used in this video (works better than the Picamera!): https://amzn.to/2MMBTU3 ---- Commands used in video ---- echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list curl packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add - wget dl.google.com/coral/canned_models/mobilenet_ssd_v2_coco_quant_postprocess_edgetpu.tflite -- Tutorial Links -- Previous TensorFlow Lite video: https://youtu.be/aimSGOAUI8Y Written version of this guide: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/blob/master/Raspberry_Pi_Guide.md Guide for training your own TFLite model: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi Performance comparison between Pi 3B+ and Pi 4 4GB: https://youtu.be/TiOKvOrYNII -- Music credit -- chill. by sakura Hz https://sakurahertz.bandcamp.com/track/chill Creative Commons — Attribution 3.0 Unported — CC BY 3.0
How To Run TensorFlow Lite on Raspberry Pi for Object Detection
10:48
Edje Electronics

How To Run TensorFlow Lite on Raspberry Pi for Object Detection

TensorFlow Lite is a framework for running lightweight machine learning models, and it's perfect for low-power devices like the Raspberry Pi! This video shows how to set up TensorFlow Lite on the Raspberry Pi for running object detection models to locate and identify objects in real-time webcam feeds, videos, or images. I used a Raspberry Pi 4 4GB for this video, but it also works with the Raspberry Pi 3. If you want to see how much faster the Pi 4 is than the Pi 3, check out my performance comparison video: https://youtu.be/TiOKvOrYNII Have questions? Ask me on Twitter @EdjeElectronics ! I usually respond faster there: https://twitter.com/EdjeElectronics -- Affiliate Links -- Get a Rasbperry Pi 4: https://amzn.to/2Kf0el8 Coral USB Accelerator: https://amzn.to/2wxTZ8d Webcam used in this video (works better than the Picamera!): https://amzn.to/2MMBTU3 -- Tutorial Links -- UPDATE (10/21/20): At 6:09 in the video, I instruct you to go to the TensorFlow Lite Object Detection Overview page and right click the "Download starter model" link to copy the link address. The page has changed since I made this video, and that link is no longer correct. Copy this link for downloading the starter model: storage.googleapis.com/download.tensorflow.org/models/tflite/coco_ssd_mobilenet_v1_1.0_quant_2018_06_29.zip Written version of this guide: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/blob/master/Raspberry_Pi_Guide.md Guide for training your own TFLite model: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi -- Music credit -- The Process by LAKEY INSPIRED: https://soundcloud.com/lakeyinspired/the-process Creative Commons — Attribution-ShareAlike 3.0 Unported — CC BY-SA 3.0
Raspberry Pi 3 vs Raspberry Pi 4 Performance with TensorFlow, TF Lite, & Coral USB Accelerator
04:45
Edje Electronics

Raspberry Pi 3 vs Raspberry Pi 4 Performance with TensorFlow, TF Lite, & Coral USB Accelerator

Have you wondered how much faster the Raspberry Pi 4 performs than the Raspberry Pi 3 at running computationally intensive TesnorFlow object detection models? This video gives a performance comparison between the Pi 3B+ and the Pi 4 4GB, showing what framerate is achieved when running TensorFlow and TensorFlow Lite SSD-MobileNet detection models. It also shows how much faster the models run when using Google's Coral USB Accelerator. This is the first video in a larger series of TensorFlow Lite videos I'm working on. The series will show how to train your own TensorFlow Lite models and run them on the Raspberry Pi, Android devices, and more. Stay tuned! Raspberry Pi 4 4GB starter kit: https://amzn.to/2Kf0el8 Coral USB Accelerator: https://amzn.to/2BuG1Tv Webcam used in this video (works better than the Picamera!): https://amzn.to/2MMBTU3 Have questions? Ask me on Twitter @EdjeElectronics ! I usually respond faster there: https://twitter.com/EdjeElectronics GitHub guide showing how to set up TensorFlow Lite and Coral USB Accelerator on the Raspberry Pi: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/blob/master/Raspberry_Pi_Guide.md GitHub guide showing how to train and convert your own TensorFlow Lite model: https://github.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi Code I used to run these tests: Will add a Dropbox link soon! ---- Music credit ---- Flamingo by jlsmrl: https://soundcloud.com/jlsmrl/free-guitar-type-beat-flamingo Creative Commons — Attribution 3.0 Unported — CC BY 3.0 Merry Bay by Ghostrifter Official: https://soundcloud.com/ghostrifter-official/merry-bay Creative Commons — Attribution-ShareAlike 3.0 Unported — CC BY-SA 3.0 Coffee Dreams by Le Gang [Audio Library Release]: https://www.youtube.com/watch?v=JAL3i9ZIwQo Music provided by Audio Library Plus
How to Set Up TensorFlow Object Detection on the Raspberry Pi
19:26
Edje Electronics

How to Set Up TensorFlow Object Detection on the Raspberry Pi

Learn how to install TensorFlow and set up the TensorFlow Object Detection API on your Raspberry Pi! These instructions will allow you to detect objects in live video streams from your Picamera or USB webcam. Get a Raspberry Pi: https://amzn.to/2Iki3fb Get a Picamera: https://amzn.to/2rKxarh Handy Picamera + Pi case: https://amzn.to/2LxaUed If you have questions, I usually respond more quickly if you send me a tweet on Twitter: @EdjeElectronics https://twitter.com/EdjeElectronics I created this video using a Raspberry Pi 3 Model B running Raspbian Stretch. It should also work for the Raspberry Pi 2. ---- Link to steps in video ---- 1:00 Step 1. Update the Raspberry Pi 1:38 Step 2. Install TensorFlow 4:14 Step 3. Install OpenCV 6:03 Step 4. Compile and Install Protobuf 10:29 Step 5. Set up TensorFlow directory structure 14:39 Step 6. Test out object detector! ---- Links mentioned in video ---- Written version of this tutorial on GitHub: https://github.com/EdjeElectronics/TensorFlow-Object-Detection-on-the-Raspberry-Pi lhelontra's "TensorFlow on ARM" repository: https://github.com/lhelontra/tensorflow-on-arm/releases OSDevLab's guide to building Protobuf on the Pi: http://osdevlab.blogspot.com/2016/03/how-to-install-google-protocol-buffers.html GitHub Protobuf releases page: https://github.com/google/protobuf/releases TensorFlow Model Zoo: https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf1_detection_zoo.md Card detector model on DropBox: https://www.dropbox.com/s/27avwicywbq68tx/card_model.zip?dl=0 Music: Broke For Free - My Always Mood http://brokeforfree.com/