How computers see and hear

A computer uses a camera to turn what it looks at into tiny dots called pixels. Each dot has a color, and millions of dots together make a picture the computer can store and look at. So the computer 'sees' by collecting lots of colored dots.

They are called pixels. A pixel is one tiny dot of color, and when many pixels sit next to each other they form a whole picture. If you look very closely at a screen, you can sometimes see them.

A computer uses a microphone to catch the sounds your voice makes in the air. The microphone turns those sounds into signals the computer can understand. That is how it can listen to what you say.

The microphone inside the smart speaker lets it hear you. It catches the sound of your words and sends them to the computer brain to understand. Then the speaker can do what you asked.

This is called face recognition. The phone's camera looks at the shape of your face and checks if it matches the picture it remembers of you. If it matches, the phone unlocks just for you.

Computers learn by seeing many examples, just like you learned animals by seeing lots of them. If a computer sees thousands of cat pictures, it starts to notice things like whiskers, ears, and fur. Then it can guess if a new picture is a cat.

It means the computer can name what is in the picture, like 'dog' or 'tree' or 'ball'. It does this by matching what it sees to patterns it learned before. Recognizing is like the computer saying, 'I know what that is!'

That is called speech-to-text. The computer listens to your voice and turns your spoken words into written words on the screen. It is like a very fast helper writing down everything you say.

A normal camera needs light to see, just like your eyes do. In the dark there is not enough light, so the picture looks black or blurry. Some special cameras use other tricks to see in the dark, but a normal one cannot.

Pixels are tiny, much smaller than a grain of rice. There are so many of them on a screen that you usually cannot see one by itself. Millions of these tiny dots work together to show a clear picture.

Sound is made of tiny waves moving through the air. A microphone catches these waves and changes them into numbers the computer can read. The computer uses those numbers to understand the sound.

Sometimes things look alike, like a brown muffin and a brown puppy face. The computer only looks at colors and shapes, so it can make a silly mistake. That is why computers are not always perfect at seeing.

The computer looks at the color of the apple in the picture. Red and green are different colors, so the computer can tell them apart. It uses the colors of the pixels to decide.

The camera on your phone or computer is sending your picture. It captures your face as pixels and sends them over the internet. The other person sees you almost right away.

Yes, a computer can look at a photo and find each face. It learned what faces look like from many examples, so it can spot them and count them. This is a kind of computer seeing.

Sometimes words sound alike, or there is loud noise around you. The computer tries its best to match the sound to a word, but it can guess wrong. That is why it sometimes does the wrong thing.

A camera's job is to capture light from the world and turn it into a picture made of pixels. It works like an eye for the computer. Without a camera, a computer could not see anything.

It might look for the road, walls, toys, or people in its way. Using its camera, it sees these things and tries not to bump into them. Seeing helps it move safely.

Your ears catch sounds from the air so you can hear. A microphone also catches sounds from the air, but it sends them to a computer. So a microphone is like an ear for the computer.

That is called text-to-speech. The computer reads the written words and says them out loud with a voice. This helps people who cannot see the words read.

A blurry photo has fuzzy, mixed-up pixels, so the shapes are not clear. The computer has trouble matching fuzzy shapes to things it knows. A clear photo is much easier for it to understand.

Yes, if it has learned the sound of barking and crying from many examples. It listens to the sound pattern and guesses which one it is. This is a kind of computer hearing.

The camera captures the view in front of it, so the computer sees whatever the camera is pointed at. If you point it at a tree, the computer sees the tree. The direction of the camera decides what it sees.

A computer with a camera can look at things and say out loud what they are, like 'a cup' or 'a door'. This helps a person who cannot see know what is around them. The computer becomes helpful eyes for them.

Adding more light makes the picture brighter and clearer. With good light, the colors and shapes show up well in the pixels. Then the computer can understand the picture more easily.

You see with your eyes and brain, but a computer only sees numbers for each pixel's color. It does not feel or understand pictures like you do. It just works with the numbers to make guesses.

It needs to see and remember a picture of your friend first. After it has learned what your friend looks like, it can recognize them later. Without learning first, it would not know who they are.

Some clever computers can guess feelings by listening to how a voice sounds, like loud and fast or soft and slow. But they are just guessing from sound patterns. They do not truly feel emotions like people do.

The computer looks for places where the color changes, like where the ball ends and the floor begins. That color change shows the edge. Finding edges helps the computer know the shape of things.

The computer stores the color of each tiny pixel as numbers. Every color, like red, blue, or green, has its own number code. All these numbers together remember the whole rainbow.

The app needs your permission to listen using the microphone. Without permission it cannot hear you, to help keep your voice private. When you say yes, it can hear you and help.

Yes, a computer can look at the shapes of the letters and figure out the words. This is like the computer reading with its camera. It learned the shapes of letters from many examples.

Either way, the computer turns the teddy bear into pixels and looks at the colors and shapes. The place does not matter, only the picture itself. It uses the same kind of seeing both times.

Both voices mix together in the air, so the sounds overlap. The computer has trouble picking out just one voice from the mix. It is easier when only one person talks at a time.