Page 21 - Ai Robogenius

2. Speed of understanding: Humans are able to process what they see fast and
frequently assign emotional significance to it. Computer Vision processes images
fast but does not understand emotions.

For example, a human can see a smiling face and interpret friendliness, whereas AI
can only identify the face as 'smiling' without understanding the context.
3. Adaptability: Human beings are able to adapt to varying lighting, angles or
new surroundings easily. Computer Vision usually needs retraining or
calibration to work in uncharted territories.

For example, a human being is able to recognise a face in low light, while an
AI may require further data to accommodate that shift.

4. Fatigue: Humans tire and may make errors when overworked. However,
Computer Vision systems never get tired and offer consistent performance.
For example, a security guard may overlook a detail after working for hours,
while an AI-based CCTV can observe around the clock without rest.

5. Learning: Human learning is shaped by experience, memory and emotion.
Children, for example, learn to recognise letters with assistance from teachers
and through practice repeated over time. Computer Vision systems learn
from labelled sets of data and feedback.

For example, AI identifies letters in scanned documents using Optical
Character Recognition (OCR).
6. Accuracy: Humans can miss the internal detail or interpret it with prejudice.
AI, if trained well, can be very accurate—but it might not grasp the bigger
picture.

For example, a doctor can miss the initial symptoms of a disease, whereas AI
can recognise patterns that suggest the symptoms, but may also send up
warning flags on innocuous cases.

7. Decision-making: Human decisions are driven by emotion, ethics and logic.
Computer Vision systems only decide based on rules and data entered.

For example, someone may allow a person to pass without an ID due to
empathy, but an AI system will block entry if the data does not align.
8. Real-life use case: Humans employ their brain and eyes to look at traffic
lights and choose when to proceed or when to brake. Self-driving cars, on
the other hand, employ Computer Vision via cameras and sensors to identify
traffic lights and determine what to do. Both are doing pretty much the same
thing, but the methods they employ are quite different.

Human vision is more intuitive and flexible, whereas computer vision is more accurate,
consistent and fast, especially when handling repetitive or large-scale tasks. However, their
complementary roles in fields such as healthcare, security and transportation can produce
remarkable results.

Computer Vision and Its Applications
19