 |
| Is Face Recognition system as effective with faces wearing during covid-19 pandemic? |
As
masks wearing scaling up because of covid-19 pandemic, companies are scrambling
to find identification based just their eyes and cheekbones.
In
current scenario, covid-19 pandemic makes mask wearing as a part of normal
life. In many places masks are used as safety precautions, and some other group
take it up as fashion trend. But this trend or requirement pose a threat on
facial recognition software performance.
The
CDC, The US Centers for Diseases and Preventions has recommended that coveringup face with Masks provide protection against fighting with covid-19 ill
caused by SARS-CoV-2 that kills more than 302,000 people in the world. In more than half of the US states, governments
have made masks wearing mandatory in various public locations.
Face
cover up with mask, and look at your iphone (for example iphone 8 face
recognition) or Android Device to unlock it, but immediately you notice the
issue of face recognition system.
Before
the unprecedented growth of novel coronavirus outbreak, the manufacturers of
face recognition systems were expecting
that it can be install at any location or place: in airports, railway stations,
casinos, restaurants, and schools.
Facial
masks use on large scale across the globe as a result of covid-19 sickness has
emerged as major threat for face recognition systems developers. On the other
hand, the industry seems more robust to handle and solve this issue as soon as
possible.
Some
companies claim that their technology, face recognition systems, isn’t affected
by mask wearing habit, and artificial intelligence is still have enough
capacity to detect and recognize people with a high accuracy rate despite of
the fact that half of the face is masked.
The
release of Apple’s latest iOS beta program reveals that big fish is engaged in
updating its Face ID (Face recognition system) to account for people having
masks on their faces. Google didn’t
speak up in response to the question of updating its android tech as did by its
competitor “Apple”.
Professionals
of face recognition technology have shown their concerns over the claim the
face recognition system isn’t deterred by masks. The situation is worst when
face recognition system stumble even if faces are free of masks. Studies have suggested that face recognition
algorithm become upset when dealing with people of color by a factor 10to 100 times. Facial recognition algorithms affecting by wearing face mask.
In
addition, due to prevailing novel covid-19 pandemic, these face recognition
systems or algorithms still not tested with face masks the US’s National
Institute of Standards and Technology or NIST, which enjoys the position of
leading authority on the facial recognition accuracy rate.
Despite
of this, face recognition software is still recommending as solution of novel
coronavirus disease in absence of any proof that support it as identifier of
covid-19 patients either with masks or without.
These
workarounds are part of the collective effort at large scale to develop an ever
surveillance or reconnaissance infrastructure a primary component of covid-19
response governance, said Evan Selinger, a professor of philosophy worked at
the Rochester Institute of Technology.
Recognizable?
Masks
have been in used since long time to hide their identity from others and
technology as well like faces recognition. Protestors in the Hong Kong relied on this method to
defeat the government’s face recognition system, prompting ban mask there.
“The
enormous amount of biometric data that uniquely include central portion of the
face starting from brow line all the way down to the chin,” said Eric Hess
Senior Product Management Director for face recognition software at SAFR, a
facial recognition company. Covering face with masks that build up a wall
between multiple significant face points and recognition algorithms that helps
to identify the person from the other.
But
still many facial recognition firms claim that their technology shows the same
performance differentiating the person despite the fact of high rate of use of
masks among the people because of covid-19 pandemic.
The
Facewatch, a UK-based company, claimed that it is releasing a technology that
uses a person’s eyes and eyebrow region as detectiveand identification points in face recognition system. The company
suggesting that its product is suitable for retail stores, and says the
development will extend to cover other factor such as religious veil called niqab
that is commonly used by Muslim women.
Other
Factors that Facewatch had already worked on person identification technology
are hats, and glasses, according to the Stuart Greenfield-company spokesman. Mostly retail stores rely on shoplifters
rather mask detection (face recognition technology), until the pandemic
emerged.
What
we demand is the government pushed the people to put up masks, and the industry
has to respond very quickly, Greedfield added. He said further, compnay’s new
technology will has the capacity to detect and identify the people using their
eyes and eyebrows focal point which don’t change over the period.
Beyond Covid-19
This
new technology could have lasting affects for long as pandemic disappears.
The
issue of public health crisis compels companies towards more accurate
identification when dealing with masks.
But later police use the same technology to identify protestors wearingface covering.
The
House Oversight Committee cautioned when face recognition will perfect, it
could have great impact on free speech and civil liberties. Police already
relied on facerecognition software but if this recognition is perfect bypassing masks, it
will create a new level of privacy for person practicing their First Amendments
Rights.
Both
Facewatch and SAFR said that they are not capable to begin work so immediately dealing
with face masks but it links with how frequently the masks are used, and rise
in demand from their customers.
There
is still room to make more development in the technology that would not have
been achieved, said Rose. “This should be an equal opportunity for all players
to scale up level of technology in terms of Surveillance.”