Talking Neural Nets

Speech synthesis is nothing new, but it has gotten better lately. It is about to get even better thanks to DeepMind’s WaveNet project. The Alphabet (or is it Google?) project uses neural networks to analyze audio data and it learns to speak by example. Unlike other text-to-speech systems, WaveNet creates sound one sample at a time and affords surprisingly human-sounding results.

Before you rush to comment “Not a hack!” you should know we are seeing projects pop up on GitHub that use the technology. For example, there is a concrete implementation by [ibab]. [Tomlepaine] has an optimized version. In addition to learning English, they successfully trained it for Mandarin and even to generate music. If you don’t want to build a system out yourself, the original paper has audio files (about midway down) comparing traditional parametric and concatenative voices with the WaveNet voices.

Another interesting project is the reverse path — teaching WaveNet to convert speech to text. Before you get too excited, though, you might want to note this quote from the read me file:

“We’ve trained this model on a single Titan X GPU during 30 hours until 20 epochs and the model stopped at 13.4 ctc loss. If you don’t have a Titan X GPU, reduce batch_size in the train.py file from 16 to 4.”

Last time we checked, you could get a Titan X for a little less than $2,000.

There is a multi-part lecture series on reinforced learning (the foundation for DeepMind). If you wanted to tackle a project yourself, that might be a good starting point (the first part appears below).

Continue reading “Talking Neural Nets”

A DIY, Visual Alexa

Talking to computers is all the rage right now. We are accustomed to using voice to communicate with each other, so that makes sense. However, there’s a distinct difference between talking to a human over a phone line and conversing face-to-face. You get a lot of visual cues in person compared to talking over a phone or radio.

Today, most voice-enabled systems are like taking to a computer over the phone. It gets the job done, but you don’t always get the most benefit. To that end, [Youness] decided to marry an OLED display to his Alexa to give visual feedback about the current state of Alexa. It is a work in progress, but you can see two incarnations of the idea in the videos below.

A Raspberry Pi provides the horsepower and the display. A Python program connects to the Alexa Voice Service (AVS) to understand what to do. AVS provides several interfaces for building voice-enabled applications:

  • Speech Recognition/Synthesis – Understand and generate speech.
  • Alerts – Deal with events such as timers or a user utterance.
  • AudioPlayer – Manages audio playback.
  • PlaybackController – Manages playback queue.
  • Speaker – Controls volume control.
  • System – Provides client information to AVS.

We’ve seen AVS used to create an Echo clone (in a retro case, though). We also recently looked at the Google speech API on the Raspberry Pi.

Continue reading “A DIY, Visual Alexa”

Raspberry Pi Want a Cracker?

If you watch the old original Star Trek, you’ll notice that the computers on board the Enterprise don’t look much like our computers (unless you count the little 3.5 inch floppies that looked pretty close to the real thing). Then again, the Enterprise didn’t need keyboards and screens since the computers did a pretty good job of listening and speaking to humans.

We aren’t quite to the point where you can just ask the computer some fuzzy open-ended question like Captain Kirk did, but we do have things like Echo, Siri, and Google Now that do a fair job of listening to you and replying. In fact, Google provides an API that can do speech recognition and generation. [Giulio] used some common Python libraries to add speech I/O to a Raspberry Pi.

Continue reading “Raspberry Pi Want a Cracker?”

Talking Star Trek

Speech generation and recognition have come a long way. It wasn’t that long ago that we were in a breakfast place and endured 30 minutes of a teenaged girl screaming “CALL JUSTIN TAYLOR!” into her phone repeatedly, with no results. Now speech on phones is good enough you might never use the keyboard unless you want privacy. Every time we ask Google or Siri a question and get an answer it makes us feel like we are living in Star Trek.

[Smcameron] probably feels the same way. He’s been working on a Star Trek-inspired bridge simulator called “Space Nerds in Space” for some time. He decided to test out the current state of Linux speech support by adding speech commands and response to it. You can see the results in the video below.

Continue reading “Talking Star Trek”

Mobile Text Reader With OCR and Text To Speech

There are devices out there that will magnify text using fancy cameras and displays, devices that will convert these to Braille, and text-to-speech software has been around for thirty years. For his entry into our Raspberry Pi Zero contest, [Markus] decided to combine all these ideas into a simple device that will turn the printed word into speech.

The impetus for [Markus]’ project came to him in the form of a group of blind computer science  students. These students used a specialized program that used specialized hardware and software such as mobile Braille terminals, OCR, and oral exams that allowed these students to study the same thing as everyone else. [Markus] wanted to produce something similar, using simple text-to-speech software instead of a complicated Braille display.

The physical design of [Markus]’ project is uniquely functional – a hand-held device with a camera up front, a Pi in the middle, and a speaker and headphone jack on the back. The hand grip includes a large battery and a trigger for telling the Pi to read a few words aloud.

The software is built around the SnapPicam and includes a lot of the functionality already needed. OCR is largely a solved problem with Tesseract, and text-to-speech is easy with Festival.

Although [Markus] is just plugging a few existing software modules together, he’s come up with a device that is certainly unique and could be exceptionally useful to anyone with a vision impairment.


Raspberry_Pi_LogoSmall

The Raspberry Pi Zero contest is presented by Hackaday and Adafruit. Prizes include Raspberry Pi Zeros from Adafruit and gift cards to The Hackaday Store!
See All the Entries

EVA: What’s on Telly for the Visually Impaired

[chewabledrapery] has certainly used his Raspberry Pi for good. His girlfriend’s grandfather is growing more visually impaired as time goes on. He likes to watch telly, but has trouble reading the on-screen information about the channel and programming. To that end, [chewabledrapery] has built an electronic voice assistant called EVA, who fetches the telly schedule from a web service and reads it aloud in her lovely voice that comes courtesy of Google Translate’s TTS function.

Under EVA’s hood is a Raspberry Pi. A USB hub powers the Pi and holds a small USB soundcard, a Wi-Fi dongle, and a USB daughterboard that the controller plugs into. The daughterboard is from a USB keyboard, which makes another appearance in the awesome controller. It’s made of a joystick and two arcade buttons that use the USB keyboard’s controller to interact with Python scripts.

[chewabledrapery]’s scripts make formatted requests to a web service called atlas, which returns JSON objects with the TV schedule and content descriptions. EVA then turns to Google Translate, speaking the formatted text through a small amplifier and salvaged PC speaker. In order to minimize the number of web calls, some of EVA’s frequent musings are stored locally. A full tour of EVA is after the break.

We love to see hacks that help people. Remember this RFID audio book reader?

Continue reading “EVA: What’s on Telly for the Visually Impaired”

Raspberry Pi Becomes a Universal Translator

hola-me-nombre-david-conroy

We’re still about 150 years away from the invention of the universal translator by [Lt Cdr Sato] of the Enterprise NX-01, but [Dave] has something that’s almost as good: a speech recognition, translation, and text to speech setup for the Raspberry Pi that theoretically allows anyone to speak in sixty different languages.

After setting up all the Linux audio cruft, [Dave] digs in and starts on converting the guttural vocalizations of a meat speaker into something Google’s speech to text service can understand. From there, it’s off to Google again, this time converting text in one language into the writings of another.

[Dave]’s end result is a shell script that works reasonably well for something that won’t be invented for another 150 years. The video below shows the script successfully translating English to spanish, but it should work equally well with other languages such as dutch and latin, as well as less popular language such as esperanto and french.

Continue reading “Raspberry Pi Becomes a Universal Translator”