There is an mistake on the website on the part where it's installing Walabot API where it states python -m pip “/usr/share/walabot/python/WalabotAPI-1.0.21.tar.gz”
#6th winfo hackathon install
Walabot plays an essential role here to make sure that someone doesn't just show a picture of you and unlock the deadbolt.įollow instructions from to install Walabot API so it can be imported to python projects. One of the biggest problem right now is that the AI can recognize your face in 2 dimension, but it does not know your face in 3 dimensional space. Step 5: Setup Walabot for person detection We can get also get the model through following command to get the caffe model in a png view python /opt/caffe/build/tools/draw_net.py /home/ubuntu/face_training/deepID_solver.prototxt /home/ubuntu/face_training/caffe_model_face.png Use train_lmdb.py that's attached in the code, you will be able to create LMDB image database that's required for training. Let's put folder under face_training just so we can get a understanding of it easily.
![6th winfo hackathon 6th winfo hackathon](https://winfo.ischool.uw.edu/img/8thHackathon.jpg)
Upload the images to the server, for start we can train 3000 of your own face image and 3000 of other people's face images.
#6th winfo hackathon free
Intel Devcloud also offers free clusters for you to train. We'd prefer use either AWS or Azure machines that are specifically built for machine learning. So for of this to work you will need a Linux operating system with either GPU or CPU installed. For this project, I've leveraged an open source project located at Specific framework we use is caffe, there are many ways to train the model but we can use some of the open source ones with right parameters. Our Alexa skill is published as Face Lock under skill id for those who are interested in taking a look In this guide we are going to convolution neural network to create facial recognition network and secure it via Walabot to detect distance as well as user breath, then open up the deadbolt via alexa. Additional radar (Walabot) will be added to the project to make sure that simple images can not pass the test. We've trained the entire network on caffe, reaching more than 99% accuracy with "me or not me" method. This project can be extended to using facial recognition to unlock deadbolts, record entries, turn on different light themes, and many others. We've built a platform that shows how the AI works on Intel Movidius NCS, using all the default camera that came with the development kit.
![6th winfo hackathon 6th winfo hackathon](https://winfo.ischool.uw.edu/img/hackathon8/allSponsorsBanner.png)
IPhone X Face ID hits people with surprise, people starting to realize through AI Deep learning their face is much more unique and accurate than their fingerprint.īut what some people don't realize is the reason why iPhone X facial recognition works is that it's only detecting you or not you, therefore it is at much higher accuracy than using AI to detect multiple target.