Categories
Research Post

Reserch Post 3

 

Image Analysis and Face Recognition

The face is presumed to be an essential element in the body of a human being, vide the fact that it is the primary feature of identifying ourselves. The Facial Action Coding System (FACS) used in determining faces is said to use forty-six dimensions of the face for recognition. The significant modifiers regarding the topic of Image analysis and face recognition include recognition, detection, and tracking, all of which connote different meanings. Detection is described as a binary classification process wherein the task involves deciding on whether something is a face or not. Recognition, on the other hand, connotes the ability to distinguish between two different faces. Lastly, tracking in so far as this topic is concerned means the liaison of both detection and recognition to help know the same face recognized prior as the same as the one recognized afterward (McDonald, 2017).

The application of facial recognition is believed to have existed for some decades. Over the past years, it has been sophisticated through various advancements to arrive at better solutions in our day-to-day lives. This includes being used for security purposes, as illustrated by the actor Denzel Washington in the movie ‘Dejavu’ (Facial Recognition: Last Week Tonight with John Oliver (HBO), 2020). The application has also since evolved to incorporate better technology used in identifying and recognizing not only human beings but also sea animals like fish.

Despite these pronounced achievements on the use of facial recognition, the issue of algorithmic bias is an issue worth concern (How I’m Fighting Bias in Algorithms | Joy Buolamwini, 2017). The computer facial recognition system is believed not to be accurate and efficient enough in recognizing faces because they all depend on the designed machine language. Consequently, this renders the use of facial recognition as being inaccurate and their legitimacy in usage such as by the security in identifying criminals may be rendered inadmissible as evidence. Through research, it is a major concern that established facial recognition software often fail to recognize black people’s faces (Richardson, 2017).

 

References

Facial Recognition: Last Week Tonight with John Oliver (HBO). (2020, June 15). YouTube. https://www.youtube.com/watch?v=jZjmlJPJgug 

How I’m fighting bias in algorithms | Joy Buolamwini. (2017, March 29). YouTube. https://www.youtube.com/watch?v=UG_X_7g63rY 

  1. (2017). Week 2 · kylemcdonald/AppropriatingNewTechnologies Wiki. GitHub. https://github.com/kylemcdonald/AppropriatingNewTechnologies/wiki/Week-2

Richardson, W. J. (2017, November 29). Against Black Inclusion in Facial Recognition. Digital Talking Drum. https://digitaltalkingdrum.com/2017/08/15/against-black-inclusion-in-facial-recognition/