3 ms·

Making a toy shark fly using brain waves

14 points by okfine 11y ago

veddox 11y ago
Wow, respect for a great hack! I wonder if there are any other known actions that give as strong and recognizable an EEG signal as closing your eyes?
- sandworm101 11y ago
  Hand in boiling water would probably fire some neurons. I don't see any wow factor here. This isn't control. It's a one-dimension (up/down) input based on activity. There are many commercially available toys that do this without any physical motion required. I've seen some at airport dutyfree. https://www.youtube.com/watch?v=RM6nwZBBbAo https://www.youtube.com/watch?v=RM6nwZBBbAo
- glia 11y ago
  Closing your eyes will produce a very large shift in the EEG spectrum, which is pretty neat, but it's a little unsatisfying since it relies on a physical movement anyhow. If you're going to move muscles on your face, EMG interference can produce signals in the same frequency bands that are an order of magnitude greater in power. Many of the popular "concentration" toys probably benefit a lot from this effect, since clenching your jaw or furrowing your brow just a bit will produce a very large change in surface potentials. Though it may be less futuristic and transcendent than decoding "pure intent", there's a lot of value for that sort of tech. Most quadriplegics retain some freedom of facial movement, and there are interfaces on the market and in development that control wheelchairs and other accessibility equipment by taking advantage of facial EMG. If you're curious about more pure or abstract EEG features that are pretty accessible to the hobbyist (not easily, but somewhat cheaply), you might be interested in the P300 wave. [0] The P300 is an event related potential. A quick example of its occurrence is after you recognize something you've been searching for. This has been used to create simple typing interfaces, in which the subject stares at a grid of letters. The event related potential shows up roughly 300 msec after their intended letter lights up. [1] Unlike the signals in the OP, ERPs tend to be less powerful than the background signal, requiring a number of averaging frames to determine their presence. This slows down the usable bandwidth enough that an able-bodied person wouldn't usually get much use out of it outside of novelty. There are, though, interesting efforts out there (some of them ML-based) to speed this up. [2] The OP discusses one of the most prominent example of synchronization and desynchronization in EEG. Similar spatio-temporal techniques can work on other regions of interest. There's been a lot of progress, for example, in decoding gross motor control from surface potentials. [3] One of the most recent advances in this vein was just discussed on HN the other day. [4][5] [0] https://en.wikipedia.org/wiki/P300_(neuroscience) https://en.wikipedia.org/wiki/P300_(neuroscience) [1] http://www.ncbi.nlm.nih.gov/pubmed/19351359 http://www.ncbi.nlm.nih.gov/pubmed/19351359 (paywall) [2] http://www1.ece.neu.edu/~erdogmus/publications/zzz_ACL2011_RSVP_Keyboard_Roark_final.pdf http://www1.ece.neu.edu/~erdogmus/publications/zzz_ACL2011_R... [3] http://www.ncbi.nlm.nih.gov/pubmed/19556679 http://www.ncbi.nlm.nih.gov/pubmed/19556679 (paywall) [4] http://www.theguardian.com/science/2015/sep/24/paraplegic-man-walks-with-own-legs-again http://www.theguardian.com/science/2015/sep/24/paraplegic-ma... [5] https://news.ycombinator.com/item?id=10275689 https://news.ycombinator.com/item?id=10275689
bazzargh 11y ago
Idea: motorize a weeping angel (https://en.wikipedia.org/wiki/Weeping_Angel https://en.wikipedia.org/wiki/Weeping_Angel), use this mechanism to have it drive towards someone when their eyes shut

14 points by okfine 11y ago

Wow, respect for a great hack! I wonder if there are any other known actions that give as strong and recognizable an EEG signal as closing your eyes?

sandworm101 11y ago
Hand in boiling water would probably fire some neurons. I don't see any wow factor here. This isn't control. It's a one-dimension (up/down) input based on activity. There are many commercially available toys that do this without any physical motion required. I've seen some at airport dutyfree. https://www.youtube.com/watch?v=RM6nwZBBbAo https://www.youtube.com/watch?v=RM6nwZBBbAo
glia 11y ago
Closing your eyes will produce a very large shift in the EEG spectrum, which is pretty neat, but it's a little unsatisfying since it relies on a physical movement anyhow. If you're going to move muscles on your face, EMG interference can produce signals in the same frequency bands that are an order of magnitude greater in power. Many of the popular "concentration" toys probably benefit a lot from this effect, since clenching your jaw or furrowing your brow just a bit will produce a very large change in surface potentials. Though it may be less futuristic and transcendent than decoding "pure intent", there's a lot of value for that sort of tech. Most quadriplegics retain some freedom of facial movement, and there are interfaces on the market and in development that control wheelchairs and other accessibility equipment by taking advantage of facial EMG. If you're curious about more pure or abstract EEG features that are pretty accessible to the hobbyist (not easily, but somewhat cheaply), you might be interested in the P300 wave. [0] The P300 is an event related potential. A quick example of its occurrence is after you recognize something you've been searching for. This has been used to create simple typing interfaces, in which the subject stares at a grid of letters. The event related potential shows up roughly 300 msec after their intended letter lights up. [1] Unlike the signals in the OP, ERPs tend to be less powerful than the background signal, requiring a number of averaging frames to determine their presence. This slows down the usable bandwidth enough that an able-bodied person wouldn't usually get much use out of it outside of novelty. There are, though, interesting efforts out there (some of them ML-based) to speed this up. [2] The OP discusses one of the most prominent example of synchronization and desynchronization in EEG. Similar spatio-temporal techniques can work on other regions of interest. There's been a lot of progress, for example, in decoding gross motor control from surface potentials. [3] One of the most recent advances in this vein was just discussed on HN the other day. [4][5] [0] https://en.wikipedia.org/wiki/P300_(neuroscience) https://en.wikipedia.org/wiki/P300_(neuroscience) [1] http://www.ncbi.nlm.nih.gov/pubmed/19351359 http://www.ncbi.nlm.nih.gov/pubmed/19351359 (paywall) [2] http://www1.ece.neu.edu/~erdogmus/publications/zzz_ACL2011_RSVP_Keyboard_Roark_final.pdf http://www1.ece.neu.edu/~erdogmus/publications/zzz_ACL2011_R... [3] http://www.ncbi.nlm.nih.gov/pubmed/19556679 http://www.ncbi.nlm.nih.gov/pubmed/19556679 (paywall) [4] http://www.theguardian.com/science/2015/sep/24/paraplegic-man-walks-with-own-legs-again http://www.theguardian.com/science/2015/sep/24/paraplegic-ma... [5] https://news.ycombinator.com/item?id=10275689 https://news.ycombinator.com/item?id=10275689

bazzargh 11y ago

Idea: motorize a weeping angel (https://en.wikipedia.org/wiki/Weeping_Angel https://en.wikipedia.org/wiki/Weeping_Angel), use this mechanism to have it drive towards someone when their eyes shut