Google’s AIY Vision Kit Augments Pi With Vision Processor

December 17, 2017 by Steven Dufresne 29 Comments

Google has announced their soon to be available Vision Kit, their next easy to assemble Artificial Intelligence Yourself (AIY) product. You’ll have to provide your own Raspberry Pi Zero W but that’s okay since what makes this special is Google’s VisionBonnet board that they do provide, basically a low power neural network accelerator board running TensorFlow.

AIY VisionBonnet with Myriad 2 (MA2450) chip

The VisionBonnet is built around the Intel® Movidius™ Myriad 2 (aka MA2450) vision processing unit (VPU) chip. See the video below for an overview of this chip, but what it allows is the rapid processing of compute-intensive neural networks. We don’t think you’d use it for training the neural nets, just for doing the inference, or in human terms, for making use of the trained neural nets. It may be worth getting the kit for this board alone to use in your own hacks. An alternative is to get Modivius’s Neural Compute Stick, which has the same chip on a USB stick for around $80, not quite double the Vision Kit’s $45 price tag.

The Vision Kit isn’t out yet so we can’t be certain of the details, but based on the hardware it looks like you’ll point the camera at something, press a button and it will speak. We’ve seen this before with this talking object recognizer on a Pi 3 (full disclosure, it was made by yours truly) but without the hardware acceleration, a single object recognition took around 10 seconds. In the vision kit we expect the recognition will be in real-time. So the Vision Kit may be much more dynamic than that. And in case it wasn’t clear, a key feature is that nothing is done on the cloud here, all processing is local.

The kit comes with three different applications: an object recognition one that can recognize up to 1000 different classes of objects, another that recognizes faces and their expressions, and a third that detects people, cats, and dogs. While you can get up to a lot of mischief with just that, you can run your own neural networks too. If you need a refresher on TensorFlow then check out our introduction. And be sure to check out the Myriad 2 VPU video below the break.

Continue reading “Google’s AIY Vision Kit Augments Pi With Vision Processor” →

Old TV Lends Case To Retro Magic Mirror

December 12, 2017 by Dan Maloney 10 Comments

Remember the days when the television was the most important appliance in the house? On at dawn for the morning news and weather, and off when Johnny Carson said goodnight, it was the indispensable portal to the larger world. Broadcast TV may have relinquished its hold on the public mind in favor of smartphones, but an information portal built into an old TV might take you back to the old days.

It seems like [MisterM] has a little bit of a thing for the retro look. Witness the wallpaper in the video after the break for proof, as well as his Google-ized Radio Shack intercom project from a few months back. His current project should fit right in, based on an 8″ black-and-white TV from the 70s as it is. TVs were bulky back then to allow for the long neck of the CRT, so he decided to lop off the majority of the case and use just the bezel for his build. An 8″ Pimoroni display sits where the old tube once lived, and replicates the original 4:3 aspect ratio. With Chromium set up in kiosk mode, the family can quickly select from a variety of news and information “channels” using the original tuning knob, while parts from a salvaged mouse turns the volume control into a scroll wheel.

It’s a nice twist on the magic mirror concept, and a little different from the other retro-TV projects we’ve seen, like a retro gaming console or an old-time case for a smart TV.

Continue reading “Old TV Lends Case To Retro Magic Mirror” →

A Wireless Webcam Without A Cumbersome Cloud Service

December 8, 2017 by James Hobson 42 Comments

After a friend bought a nannycam that required the use of a cloud service to make the device useful, [Martin Caarels] thought to himself — as he puts it — ”I can probably do this with a Raspberry Pi!”

Altogether, [Caarels] gathered together a 4000mAh battery, a Raspberry Pi 3 with a micro SD card for storage, a Logitech c270 webcam, and the critical component to bind this project together: an elastic band. Once he had downloaded and set up Raspbian Stretch Lite on the SD card, he popped it into the Pi and connected it to the network via a cable. From there, he had to ssh into the Pi to get its IP so he could have it hop onto the WiFi.

Now that he effectively had a wireless webcam, it was time to turn it into a proper security camera.

Continue reading “A Wireless Webcam Without A Cumbersome Cloud Service” →

Woman Gets Diabetes, Builds Own Pancreas

December 6, 2017 by Will Sweatman 64 Comments

For the most part, when we break out the soldering iron to make a project for ourselves – we do so for fun. Sometimes we do so for necessity. Rarely do we, however, do so to save our own lives. [Dana Lewis] is one of the 30 million people in the US who suffer from diabetes. It’s a condition where the pancreas fails to make insulin, resulting in a buildup of sugar in the bloodstream. Managing the levels of insulin and sugar in their bodies is a day-to-day struggle for the millions of diabetics in the world. It’s a great deal more for [Dana], however. She sleeps with machines that monitor the glucose levels in her blood, but lives with constant worry.

“I was afraid at night because I am a super-deep, champion sleeper,” Lewis said, “I sleep through the alarms on the device that are supposed to wake me up and save my life…”

What she needed was the glucose data from the device and use it to trigger a louder alarm. It wasn’t long until she found someone who had done just this. Using a Raspberry Pi, she was able to capture the data and then alarm her via her phone. She then setup a web interface so others could see her data and call her if she didn’t wake.

The next step is obvious. Why not make the state of the insulin pump a function of the data? And thus, a sort of artificial pancreas.

The project is open source for anyone to use and improve upon. She was placed on a list for the 100 most creative people in the US for 2017. We’re not strangers to the idea of an artificial pancreas, but it’s always great to see people using things we make video game consoles out of to save lives.

Thanks to [Dave Zzzz] for the tip!

Astro Cat: Raspberry Pi Telescope Controller

December 5, 2017 by Tom Nardi 8 Comments

When somebody tackles an engineering problem, there are two possible paths: they can throw together a quick and dirty fix that fits their needs (the classic “hack”, as it were), or they can go the extra mile to develop a well documented solution that helps the community as a whole. We cover it all here at Hackaday, but we’ve certainly got a soft spot for the latter approach, even if some may feel it falls into the dreaded territory of “Not A Hack”.

When [Gary Preston] wanted to control his telescope and astrophotography hardware, he took the second path in a big way. Over the course of several posts on his blog, [Gary] walks us though the creation of his open source Raspberry Pi add-on board that controls a laundry list of sensors and optical gear. Just don’t call it a HAT, while it may look the part, [Gary] is very specific that it does not officially meet the HAT specifications put out by the Raspberry Pi Foundation.

Even if you aren’t terribly interested in peering into the infinite void above, the extremely detailed write-up [Gary] has done contains tons of multidisciplinary information that you may find useful. From showing how to modify the Pi’s boot configuration to enable true hardware UART (by default, the Pi 3 ties it up with Bluetooth) and level shifting it with a ST3232 to a breakdown of the mistakes he made in his PCB layout, there’s plenty to learn.

Astro CAT is a completely open source project, with the hardware side released under the CERN Open Hardware License v1.2, and the INDI driver component is available under the GPL v3.

If this looks a bit daunting for your first stab at astrophotography with the Raspberry Pi, fear not. We’ve covered builds which can get you up and running no matter what your budget or experience level is.

Meet The Modern Meat Man’s Modified Meat-Safe

December 3, 2017 by Lauren Faris 5 Comments

Charcuterie is delicious — but is it hackable? When talking about the salty preserved meats, one might be more inclined to indulge in the concept of bacon before pondering a way to integrate an electrical monitoring system into the process. However, [Danzetto] decided to do both when he did not have anywhere to cure his meats. He made his own fully automatic meat curing chamber lovingly called the curebOS with the aid of a raspberry pi. It is basically a beefed up mini fridge with all of the bells and whistles.

This baby has everything. Sitting on top is a control system containing the Pi. There are 5 relays used for the lights, circulating fan, ventilating fans, refrigerator, and humidifier all powered by a 5 amp supply — minus the fridge. Down below that is the 3D printed cover with a damper for one of the many ventilation fans that regulate the internal temperature. To the right is a touchscreen for viewing and potentially controlling the system if necessary. The control program was written in Python for viewing the different trends. And below that, of course, is a viewing window. On the inside are temperature and humidity probes that can be monitored from the front screen. These readings help determine when to activate the compressor, any of the fans, or the humidifier for optimal settings. For a final touch, there are also some LEDs placed above the hanging meat to cast a glowing effect upon the prized possessions.

Continue reading “Meet The Modern Meat Man’s Modified Meat-Safe” →

Purely Functional Selfies: Thermal Printer Speaks Haskell

December 1, 2017 by Kristina Panos 6 Comments

[Dan] recently got a cheap POS thermal printer to chooch remotely over ESP32. Having conquered that project, he decided to see what else he could get the printer to do. Why not use it to print pictures? Sure, it’s been done, but not with Haskell. And yeah, the pictures will be grainy and weird-ish and limited to black and white, but hey, we love black and white around here as much as the idea of doing something simply because you can.

In the first project, [Dan] had to figure out how to talk to the printer since the RS422 cable it came with didn’t seem to work. He bought a TTL-to-RS485 adapter, but then realized he could use TTL directly and wired up a ESP32/OLED dev board to it. During the course of turning it into a photo booth, he had to switch to a bigger screen with a better refresh rate.

Unfortunately, [Dan] was unable to use Haskell by itself. He blames this on the cobwebs in the Haskell ecosystem, something that isn’t a problem for languages like Python that celebrate wide usage and support. [Dan] wrote a Python script that handles image capturing, display, and listening for touch activity on the screen, but Haskell ultimately controls the printer. Check out [Dan]’s demo after the break.

This project may have been trying at times, but at least [Dan] didn’t have to give it a brain transplant to get it to do what he wanted.

Continue reading “Purely Functional Selfies: Thermal Printer Speaks Haskell” →