Depth estimation basics in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is depth estimation in computer vision?

Depth estimation is the process of determining the distance of objects from the camera or observer in an image or video.

beginner

Name two common methods used for depth estimation.

Two common methods are stereo vision (using two cameras) and monocular depth estimation (using one camera with AI models).

intermediate

Why is stereo vision effective for depth estimation?

Stereo vision uses two cameras to capture images from slightly different angles, allowing the system to calculate depth by comparing differences between the images, similar to how human eyes work.

intermediate

What role do AI models play in monocular depth estimation?

AI models learn patterns from many images with known depths to predict depth from a single image, even without multiple camera views.

beginner

What is a common output format of depth estimation models?

The output is usually a depth map, which is a grayscale image where each pixel's brightness represents the distance of that point from the camera.

What does a depth map represent?

A Color of objects in the image

B Speed of moving objects

C Distance of objects from the camera

D Texture details of objects

A depth map shows how far each point in the image is from the camera.

Which method uses two cameras to estimate depth?

A Stereo vision

B Semantic segmentation

C Monocular depth estimation

D Image classification

Stereo vision uses two cameras to compare images and calculate depth.

Why is monocular depth estimation challenging?

A Because it uses two cameras

B Because it only works outdoors

C Because it needs 3D sensors

D Because it requires AI to guess depth from one image

Monocular depth estimation must predict depth from a single image, which is harder and needs AI learning.

What is the main input for depth estimation?

A Images or videos

B Audio signals

C Text data

D Temperature readings

Depth estimation uses images or videos as input to find distances.

Which of these is NOT a use of depth estimation?

A 3D modeling

B Speech recognition

C Robot navigation

D Augmented reality

Speech recognition does not use depth estimation; it processes audio.

Explain how stereo vision helps in estimating depth.

Think about how your two eyes help you see depth.

Describe the challenges and solutions of monocular depth estimation.

How can AI guess depth from just one photo?