Skip to content

advait-l/Eye-Blink-Detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Eye Blink Detection

Source Encoding Model

  • At present, we are using 10 bits to encode the data bit. Thus, we can think of this as a Repetition(10, 1) scheme.

  • We also are using all 16 combinations of the 4 bit input data sequence (40 bit encoded sequence). Thus, the input data has a minimum hamming distance of 1

  • If we combine this fact along with our encoding scheme, the effective minimum hamming distance is 10

    formula

  • Thus, this encoding scheme can correct upto 4 bits.

  • Similarly, if we increase the minimum hamming distance of the input data to 2 by considering only 4 out of the 16 4 bit schemes, the minimum effective hamming distance is now 20, and this can correct upto 9 bits.

Probability of error of transmission

  • The probability of error of transmission is estimated by taking trials against the ground truth blink sequence 10101010101010101010.
  • Each bit of the ground truth sequence has a corresponding 10 bit of sensor data. So the ground truth blink sequence corresponds to a 100-bit binary sequence 111111111100000000001111111111...
  • This sequence was chosen as the ground truth because this sequence capture the transition of the bits from 0 to 1 and 1 to 0.
  • Multiple trials of sequences are conducted against the ground truth blink sequence.
  • The probability of transmission error was calculated as the hamming distance between the recorded sequence and the ground truth sequence.
  • It was observed that the probability of transmission error ranged from 0.1 to 0.4 for various trials. The mean probability comes out to be 0.28

Theoretical Evaluation

  • The probability of error for an encoding scheme can be theoretically calculated with the help of the number of bits that can be corrected.

formula

img1 Input vs Output probability of error img2 Log plot for the same

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages