Brain Computer Interface Done!!

@rupam-das-wW5CvN • Oct 26, 2024

Oct 26, 2024

1.3K

Today We succeed in Brain computer Interfacing at our Lab. We used 6 serially connected Electrode ( dry) to acquire EMG ( Brain Signal). We than Extracted the features after post processing and Trained a Neural Network at our various thought process.

The signal acquired from Brain is Processed in PC which controls an External circuit of Bulb.

Just by thinking that the bulb is getting dimmed, we are able to dim it increase the intensity. Also Switch on and Off are achieved.
67% accuracy, 32 seconds response time in average. and a dam low cost one with only $50.

Now the technology will be patented with some more result improvement. Feeling fantastic happy and feeling totally drained out after nearly 3 months of relentless efforts.

Gotta go to sleep............................ bye Guys!

Like 0 Replies 14

Replies

Welcome, guest

Join CrazyEngineers to reply, ask questions, and participate in conversations.

CrazyEngineers powered by Jatra Community Platform

Ankita Katdare

@abrakadabra • Dec 28, 2011

This is awesome Rupam! Pathbreaking. 👍
We would like to know more about this from you. Do share the video too.

0
Harshad Italiya

@harshad-ukH5ww • Dec 28, 2011

Wow ! Superb Work there Rupam, waiting for some more details. 😀

0
ISHAN TOPRE

@ishan-nohePN • Dec 28, 2011

This calls for a video and most importantly a PARTY 😀 Awesome work friend 👍

0
Kaustubh Katdare

@thebigk • Dec 28, 2011

Can't wait to see the photos and videos of the wizardry you've done 😉 Thread moved to electronics section.

0
Harshad Italiya

@harshad-ukH5ww • Dec 28, 2011

Signals you are receiving from sensors connected to brain is remains uniform for all humans or it may vary according to person?

0
Rupam Das

@rupam-das-wW5CvN • Jan 4, 2012

Hello all, First of all thank you all for your posts. 😔 Unfortunately due to some real hard deadline, I am unable to follow CE threads, But here is what is the summery:
1) There are total 128 Points in brain where neurons meets and where the signal is maximum.
2) We identified 8 such points where the signal strength is maximum
3) A Cap was designed with points placed at those points ( They may vary from person to person but will be almost at the same position)
4) The anode of one point was connected to the cathode of the other.
5) Finally we took the actual anode and cathode. ( the signal range is 2mV to 27mV)
6) That signal is amplified using Instrumentation amplifier
7) The impedance matching was performed by adding a buffer at the output stage.
8) Signal was read to PC by ATMEG 168 through USB Shield.
9) This signal is Non stationary.
10) It is made stationary with STFT.
11) Features were extracted from the signal ( total 128 features)
12) We trained a Neural Network with 5000 epochs and .00001 Error rate with 512 Hidden layers. for 4 Input conditions.

Bingo.........

0
Rupam Das

@rupam-das-wW5CvN • Jan 12, 2012

Update: Video of Signal Processing part in BCI

0
Dancer_Engineer

@dancer-engineer-EJ8rGI • Jan 12, 2012

Awesome Rupam. 👍 😀
Rupam Sir. 😀

0
Kaustubh Katdare

@thebigk • Jan 12, 2012

Impressive, Rupam! 😀

0
Saandeep Sreerambatla

@saandeep-sreerambatla-hWHU1M • Jan 21, 2012

Very good work Rupam. Is this related to the 8 points in lie detector? Curious to know. Thanks!

0
lovejeet

@lovejeet-etHdkD • Jan 21, 2012

really awesome work. hats off to the awesome effort you all put in to the project. congrats.

0
PraveenKumar Purushothaman

@praveenkumar-66Ze92 • Jan 21, 2012

Looks cool buddy... Where can we see the demo?

0
ashuashi

@ashuashi-Z4179a • Jan 24, 2012

@ Rupam... Gud Job...

Once fully done and tested, can be further improved to wireless transmission... 👍

0
Irfan Amin

@irfan-amin-ro8gwr • Sep 22, 2016

Rupam DasUpdate: Video of Signal Processing part in BCI

can you help me in this project. I'm currently working on this project

0