As a fun project I thought I’d put Google’s Inception-v3 neural network on a Raspberry Pi to see how well it does at recognizing objects first hand. It turned out to be not only fun to implement, but also the way I’d implemented it ended up making for loads of fun for everyone I showed it to, mostly folks at hackerspaces and such gatherings. And yes, some of it bordering on pornographic — cheeky hackers.
An added bonus many pointed out is that, once installed, no internet access is required. This is state-of-the-art, standalone object recognition with no big brother knowing what you’ve been up to, unlike with that nosey Alexa.
But will it lead to widespread useful AI? If a neural network can recognize every object around it, will that lead to human-like skills? Read on. Continue reading “DIY Raspberry Neural Network Sees All, Recognizes Some”
[HydroGraphix HeadQuarters] has earned his name with this one. While he is using mineral oil instead of hydro, he’s certainly done a nice job with the graphics of it. The ‘it’ in questions is an overclocked Raspberry Pi 3 in a transparent container filled with mineral oil, and with a circulating fan.
He’s had no problem running the Pi at 1.45 GHz while running a Nintendo 64 emulator, getting between 40 °C and 50 °C. The circulating fan is a five volt computer USB fan. It’s hard to tell if the oil is actually moving, but we’re pretty sure we see some doing so near the end of the video below the break.
Mineral oil is not electrically conductive, and is often used to prevent arcing between components on high voltage multiplier boards, but those components are always soldered together. If you’ve ever worked with mineral oil, you know that it creeps into every nook and cranny, making us wonder if it might work its way between some of the (non-soldered) contacts in the various USB connectors on this Raspberry Pi. Probably not, but those of us with experience with it can attest to it’s insidiousness.
Continue reading “Liquid Cooling Overclocked Raspberry Pi With Style”
Eager to get deeper into robotics after dipping my toe in the water with my BB-8 droid, I purchased a Raspberry Pi 3 Model B. The first step was to connect to it. But while it has built-in 802.11n wireless, I at first didn’t have a wireless access point, though I eventually did get one. That meant I went through different ways of finding it and connecting to it with my desktop computer. Surely there are others seeking to do the same so let’s take a look at the secret incantations used to connect a Pi to a computer directly, and indirectly.
Continue reading “Networking: Pin the Tail on the Headless Raspberry Pi”
I had great fun writing neural network software in the 90s, and I have been anxious to try creating some using TensorFlow.
Google’s machine intelligence framework is the new hotness right now. And when TensorFlow became installable on the Raspberry Pi, working with it became very easy to do. In a short time I made a neural network that counts in binary. So I thought I’d pass on what I’ve learned so far. Hopefully this makes it easier for anyone else who wants to try it, or for anyone who just wants some insight into neural networks.
Continue reading “Introduction To TensorFlow”
Looking for something a bit more from your Raspberry Pi? Tired of the usual console and arcade games? Eltech’s Exagear Desktop is a virtual machine that runs on your Raspberry Pi and allows you to run x86 games. [Dmitry]’s done a write-up about running more modern games on your Raspberry Pi.
Up until now, the Pi has been a great platform for retro gaming. By running MAME or EmulationStation, you can play classic arcade games as well as the great console games you played as a kid. Exagear Desktop goes one further, allowing you to use Wine to play more modern PC games on your Raspberry Pi 3.
The Pi 3 is still a bit underpowered for bleeding edge games, but is powerful enough that it can play some of the PC games from a few years ago. [Dmitry]’s example shows how to get Arcanum, Disciples II, and Fallout running on the Raspberry Pi. In the second part of the write-up, [Dmitry] shows you how to get Heroes of Might and Magic 3, Sid Meier’s Alpha Centauri, and Caesar 3 installed and running as well.
Obviously they will always lag behind today’s gaming machines, but the power now available in a computer the size of a credit card is pretty impressive. It’s nice to have a tool that allows one to play more than just the console games from years gone by — this opens up a whole range of great PC games to add to our library. Maybe it’s time to fabricate that new PC game controller. Or, if the Raspberry Pi seems like too much power, you could consider playing retro games on an Arduino.
[Dan Englender] was working on implementing a home automation and security system, and while his house was teeming with sensors, they used a proprietary protocol which was not supported by the open source system he was trying to implement. The problem with home automation and security systems is the lack of standardization – or rather, the large number of (often incompatible) standards used to ensure consumers get tied in to one specific system. He has shared the result of his efforts at getting the two to talk to each other via his project decode345.
The result enabled him to receive signals from Honeywell’s 5800 series of wireless products and interface them with OpenHAB — a vendor and technology agnostic open source automation software. OpenHAB offers “bindings” that allow a wide variety of systems and hardware to be integrated. Unfortunately for [Dan], this exhaustive list does not yet include support for the (not very popular) 345MHz protocol used by the Honeywell 5800 system, hence his project. Continue reading “Using SDR to Take Control of Your Home Security System”
A rumor that has been swirling around the Raspberry Pi hardware community for a significant time has proven to have a basis in fact. The Raspberry Pi 2 has lost its BCM2836 32-bit processor, and gained the 64-bit BCM2837 processor from its newer sibling, the Raspberry Pi 3. It seems this switch was made weeks ago without any fanfare on the release of the Pi 2 V1.2 board revision, so we are among many news sources that were caught on the hop.
The new board is not quite a Pi 3 masquerading as a Pi 2 though. The more capable processor is clocked at a sedate 900MHz as opposed to the Pi 3’s 1.2GHz and there is no Bluetooth or WiFi on board, but the new revision will of course benefit from the extra onboard cache and the 64-bit cores.
This move almost certainly has its roots in saving the cost of BCM2836 production in the face of falling Pi 2 sales after the launch of the Pi 3. It makes sense for the Foundation to keep the Pi 2 in their range though as the board has found a home in many embedded products for which the Pi 3’s wireless capabilities and extra power consumption are not an asset.
Avid collectors of Pi boards will no doubt be running to add this one to their displays, but given that the Pi 2 sells for the same price as a Pi 3 we suspect that most Hackaday readers will go for the faster board. It is still a development worth knowing about though, should you require a faster Pi that is a little less power-hungry. The full specification for the revised board can be found on the Raspberry Pi web site.
The Pi has come a long way since the morning in 2012 when our community brought down the RS and Farnell websites trying to buy one of the first models. This BCM2837 board joins a BCM2837-powered Compute Module as well as the Pi 3. It’s worth reminding you though that there are other players to consider, earlier this year we brought you a look at the Odroid C2, and of course the infamous Apple Device.
Pi 2 header image: Multicherry [CC BY-SA 4.0], via Wikimedia Commons.
Editorial Note: We originally covered this in Sunday’s Links article but thought it warranted another, expanded mention.