State-of-the-art Real-time Action Recognition
This is a fork of the TwentyBn:Sense repository where I'll be building a few more demos with Sense.
For installation instructions and instructions on running the already included demos, check out the readme on the main repository. You'll have to also download the pre-trained weights as described there too.
The full video can be found here.
In this fork a "RisottoStone" demo is added which is a cooking assistant that can detect some basic kitchen activities. For more details, check out my blogpost on it here.
To run RisottoStone:
python examples/run_risotto_stone.py --use_gpu
Note: This example uses a basic Text-to-speech tool called spd-say which is pre-installed in Ubuntu 14.04+. On
older versions or on MacOS, you'll have to install it manually using sudo apt install speech-dispatcher
The code is copyright (c) 2020 Twenty Billion Neurons GmbH under an MIT Licence. See the file LICENSE for details. Note that this license only covers the source code of this repo. Pretrained weights come with a separate license available here.
The code makes use of these sounds from freesound:
- "countdown_sound.wav" from "milton." licensed under CC0 1.0
- "done_sound.wav" and "exit_sound.wav" from "paep3nguin" licensed under CC0 1.0