The covid lock-down is sentencing a lot of people to work from home. For that we need webcams, a new pc, better monitor  and other equipment. The Raspberry Pi 400 is  a great and affordable second desktop, it’s the fastest Pi available suited for desktop use. It lacks a camera connector, so you need a Zero to connect a camera board.

So what about reusing that old gadget that was collecting dust in the drawer: a small and cheap Raspberry Pi Zero. If you also happen to own a camera board for that little computer, here is a nice project showmewebcam to turn that cheap computer into a handy webcam. It uses a small Buildroot Linux version and boots really fast.

And now that project has been forked on Github: showmewebcam-usbboot. No SD-card needed. That project will let the Raspberry Pi Zero boot over USB!

If you don’t have a camera board. You can buy a clone version of the v1 board for a few dollars in China, or get a v2 board or for the best results the really great HQ-camera board.

Interested in photography, that High Quality camera board can do some really great macro-photography.

Cheap Raspberry Pi webcam

With the Raspberry Pi Zero costing about 5 euro, and a clone camera board around 3 euro on AliExpress, you can have a very nice and decent webcam for less then 10 euro’s. And that is a bargain. Remember you don’t need an SD card. So save on the hardware where you can.

How does it work?

How does the Pi boot then? Over USB, so you do need the usbboot/rpiboot tool. That is free and opensource software made by the Raspberry Pi organization, a program that makes the Raspberry P Zero boot by pushing the operating system over USB.

Let’s try it out. I’m using an Ubuntu desktop.

Step 1: download showmewebcam-usbboot

Download the latest release .

Unzip it. You will extract a directory called showmewebcam-usbboot.

Step 2: install usbboot/rpiboot

If you have already installed a version of this tool, you can of course skip this tool and jump to step 3.

Download the usbboot/rpiboot, unzip it and build it according to the instructions.

Step 3: Connect the raspberry Pi Zero

(with installed camera board) by putting the USB plug into the middle USB connector.

Step 4: boot the webcam

You can start the webcam by executing the rpiboot tool and pointing it to the unzipped showmewebcam-usbboot directory

sudo ~/usbboot/rpiboot -d ~/showmewebcam-usbboot

After about half a minute booting the Raspberry Pi Webcam will be ready.

How to use the Webcam on the PC

For a quick start, and when no other cam is connected, you can start the webcam with (Ubuntu):

mpv /dev/video0

On a laptop with a build-in webcam the command that you have to issue will be :

mpv /dev/video2

To control your cams on the PC, you need `v4l-utils` (apt install v4l-utils)

To set the resolution

v4l2-ctl --set-fmt-video=width=1280,height=720 -d /dev/video2

To list all video devices:

v4l2-ctl --list-devices

To list the specs of your video devices:

v4l2-ctl --list-formats-ext

ffmpeg -f v4l2 -list_formats all -i /dev/video2

A serial connection will also be available, so you can connect to the Webcam to control the settings and do debugging. That works exactly the same as in showmewebcam

Control the webcam

Finer control offers the camera-ctl tool on the Zero

Connect to the Zero:

sudo screen /dev/ttyACM0 115200 Start the tool # /usr/bin/camera-ctl

Controlling the webcam with camera-ctl

Create a stereo webcam

The tool will let you boot more then one Raspberry Pi Zero from the same directory. Just connect the two Raspberry Pi Webcams and start the tool. One will boot, just execute the tool another time and the other one will boot.

The webcams will be available as /dev/video0 (you) and /dev/video2 (your cat) on your Pi400.

Now make that video call with your boss, and ask for a raise. Switch the webcam to your cat the moment you stopped speaking.

What can showmewebcam-usbboot do what showmewebcam can’t do?

Nothing. It can do the same with less hardware, because it doesn’t need an SD card. But is does need extra software, which can limit compatibility. It also boots slower.

It hasn’t been tested that much.

Have fun, try it out,  en let me now your thoughts!

Links:

• https://github.com/Janghou/showmewebcam-usbboot
• https://github.com/showmewebcam/showmewebcam/
• https://github.com/raspberrypi/usbboot
• https://github.com/peterbay/uvc-gadget

Deep Speech is an open speech-to-text engine by Mozilla. Speech synthesis and Speech to text are fun to try out, and I read that it could run on a Raspberry Pi4 with ease on one core, so I decided to give it a try.

The Raspberry Pi version is using Google’s TensorFlow Lite for an implementation of Baidu’s DeepSpeech architecture.

Installing it on a Raspberry 4 Buster distribution was not straightforward. First I read instructions on the Github page and tried to download and install the git version and, but I ran into problems. It was taking ages and I ran into the famous `wheels` problem.

Failed building wheel for scipy

After tweaking and trying a few times, i gave up on the Github version and tried the instructions here, but also that was a bumpy road. But success waits in the end.

Let’s go, how to install DeepSpeech on the RPI4

Create a dev directory:

mkdir dev
cd dev

Create a Python Virtual environment.

python3 -m venv deepspeech-train-venv

Activate the virtual environment

source dev/deepspeech-train-venv/bin/activate

Create the deepspeech directory

mkdir deepspeech
cd deepspeech

Install deepspeech

pip install deepspeech

Download pre-trained English model

curl -LO https://github.com/mozilla/DeepSpeech/releases/download/v0.6.0/deepspeech-0.6.0-models.tar.gz
tar xvf deepspeech-0.6.0-models.tar.gz

Download example audio files

curl -LO https://github.com/mozilla/DeepSpeech/releases/download/v0.6.0/audio-0.6.0.tar.gz
tar xvf audio-0.6.0.tar.gz

Done, run, well , eh, I tried to run the example on the instruction page

deepspeech --model deepspeech-0.6.0-models/output_graph.pbmm --lm deepspeech-0.6.0-models/lm.binary --trie deepspeech-0.6.0-models/trie --audio audio/2830-3980-0043.wav

Errors!?!  I installed a missing dependency:

sudo apt install libatlas3-base

Still errors

ModuleNotFoundError: No module named 'numpy.core._multiarray_umath'

So I check if I had numpy installed

pip install numpy
Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple
Requirement already satisfied: numpy in /home/pi/dev/deepspeech-train-venv/lib/python3.7/site-packages (1.15.4)

I decided to update numpy:

pip install --upgrade numpy
Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple
Collecting numpy
Using cached https://www.piwheels.org/simple/numpy/numpy-1.18.0-cp37-cp37m-linux_armv7l.whl
tensorboard 2.0.2 has requirement setuptools>=41.0.0, but you'll have setuptools 40.8.0 which is incompatible.
Installing collected packages: numpy
Found existing installation: numpy 1.15.4
Uninstalling numpy-1.15.4:
Successfully uninstalled numpy-1.15.4
Successfully installed numpy-1.18.0

So i decided to update setuptools too:

pip install --upgrade setuptools
Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple
Collecting setuptools
Using cached https://files.pythonhosted.org/packages/f9/d3/955738b20d3832dfa3cd3d9b07e29a8162edb480bf988332f5e6e48ca444/setuptools-44.0.0-py2.py3-none-any.whl
Installing collected packages: setuptools
Found existing installation: setuptools 40.8.0
Uninstalling setuptools-40.8.0:
Successfully uninstalled setuptools-40.8.0
Successfully installed setuptools-44.0.0

I tried to run the example on the instruction page again
# Transcribe an audio file

deepspeech --model deepspeech-0.6.0-models/output_graph.pbmm --lm deepspeech-0.6.0-models/lm.binary --trie deepspeech-0.6.0-models/trie --audio audio/2830-3980-0043.wav

Another error

Loading model from file deepspeech-0.6.0-models/output_graph.pbmm
TensorFlow: v1.14.0-21-ge77504a
DeepSpeech: v0.6.0-0-g6d43e21
ERROR: Model provided has model identifier '='+;', should be 'TFL3'

Didn’t work. I needed to change the model to `tflite`

deepspeech --model deepspeech-0.6.0-models/output_graph.tflite --lm deepspeech-0.6.0-models/lm.binary --trie deepspeech-0.6.0-models/trie --audio audio/2830-3980-0043.wav

Success in the end!

Loading model from file deepspeech-0.6.0-models/output_graph.tflite
TensorFlow: v1.14.0-21-ge77504a
DeepSpeech: v0.6.0-0-g6d43e21
INFO: Initialized TensorFlow Lite runtime.
Loaded model in 0.0019s.
Running inference.
why should one hault on the way
Inference took 4.091s for 2.735s audio file.

Then I played the audio-file:

aplay audio/4507-16021-0012.wav

Must say DeepSpeech is much smarter then me, I couldn’t understand it:
why should one hault on the way

BTW good question. No I need another engine to answer that!

Way to go, folks.