Top 30 Easy Computer Vision Concepts Quiz for Beginners
30 easy multiple-choice Computer Vision concept questions, the vocabulary and first principles, the parts you need before anything else makes sense. They come from 30 bites in the Computer Vision library, the gentlest slice of the 166 Computer Vision concept questions in the library. Answer them here or read straight down. Every question carries the correct option, why it is correct, and a link to the bite it came from.
Image/video models, diffusion, OCR, multimodal
30 questions. Pick an answer, or open “Show the answer” to read it.
Answers are graded in your browser. Nothing is saved, and no XP or streak is earned here. The app keeps score.
Question 1 of 30
What is the main drawback of representing a digital image as a fixed grid of pixels?
Show the answer
Answer: b · It causes the image to appear blurry or pixelated when scaled up significantly.
The card states that the 'primary weakness is scaling' for raster images, as making them larger requires inventing new pixels, leading to blurriness or pixelation. Option D describes vector graphics, which is a common misconception about how raster images work.
Question 2 of 30
What is the primary reason the pinhole camera model should not be directly applied to raw images from real cameras?
Show the answer
Answer: d · It does not model the non-linear distortions introduced by real camera lenses.
The card explicitly states that real cameras introduce non-linear distortions (e.g., barrel or fisheye effects) that violate the straight-line assumption of the pinhole model. While other issues might exist, lens distortion is the primary limitation mentioned for direct application.
Read the full bite: Pinhole Camera Model: Projecting 3D to 2D
Question 3 of 30
In the RGB color model, what color is produced when red, green, and blue light are combined at their maximum intensity?
Show the answer
Answer: a · White
The RGB model is additive, meaning it starts with black and adds light. The card explicitly states that when all three primary lights (red, green, blue) overlap at full intensity, the result is white. Black is the result of mixing all primary pigments in a subtractive model, which is a common misconception the card addresses.
Read the full bite: RGB Color Model: Mixing Light, Not Paint
Question 4 of 30
Which of the following best describes a fundamental limitation of an image histogram?
Show the answer
Answer: d · It fails to provide any information about the spatial arrangement of pixels in an image.
An image histogram has zero spatial awareness; it summarizes pixel counts by brightness but does not indicate where those pixels are located in the image. Therefore, it cannot judge composition. The other options describe capabilities that histograms possess or misrepresent their core function.
Read the full bite: Image Histograms: Visualizing an Image's Tonal DNA
Question 5 of 30
For many computer vision tasks, what is the main benefit of converting a color image to grayscale?
Show the answer
Answer: d · It reduces the computational complexity by representing each pixel with a single brightness value, aiding tasks focused on shape or texture.
Grayscale conversion simplifies image data by reducing each pixel from three color values (RGB) to a single brightness value, making algorithms faster and easier to develop for tasks that rely on shape, texture, or contrast. Option C is incorrect because grayscale's primary purpose in computer vision is data simplification for machine processing, not enhancing visual clarity for human interpretation.
Read the full bite: Grayscale Conversion: Seeing in Shades of Gray
Question 6 of 30
To ensure optimal image quality when an asset is needed at multiple display sizes, what is the recommended approach?
Show the answer
Answer: c · Store only the highest resolution version and generate each specific size from that original.
The card states, "For assets needed at multiple sizes, always generate each version from the high-resolution original." This prevents the cumulative quality loss that occurs when an image is scaled multiple times, especially when downscaled and then upscaled, as data is permanently discarded with each downscaling operation.
Read the full bite: Image Scaling: Resizing Pixels Without Ruining Them
Question 7 of 30
What is a significant drawback of applying histogram equalization to an image?
Show the answer
Answer: c · It can dramatically amplify existing noise, especially in dark regions.
The card explicitly states that histogram equalization's "biggest weakness is that it can dramatically amplify noise." Option D is incorrect because the process is described as an "automatic method" that calculates the cumulative distribution function.
Read the full bite: Histogram Equalization: Spreading Out Pixel Brightness
Question 8 of 30
What is the main drawback of using the Harris Corner Detector in computer vision applications?
Show the answer
Answer: b · Its performance is not robust to variations in image scale or resolution.
The card explicitly states that a "key weakness is its sensitivity to image scale" and advises against its use when an application "must handle images at different sizes or zoom levels." Other options are either contradicted by the text (e.g., it's described as fast and effective) or not highlighted as its primary limitation.
Read the full bite: Harris Corner Detector: Finding Sharp Changes in Images
Question 9 of 30
What unique characteristic makes a point suitable for detection by a corner detection algorithm?
Show the answer
Answer: d · Shifting a small window over it in any direction causes a noticeable change in pixel intensity.
The card explains that a corner is identified because moving a small window over it in any direction causes a significant change in what is seen. Option B describes an edge, where intensity changes sharply in only one direction, not multiple.
Read the full bite: Corner Detection: Finding Stable Points in Images
Question 10 of 30
What is the fundamental mechanism the Hough Transform uses to detect shapes in an image?
Show the answer
Answer: d · It identifies local maxima in a parameter space where edge pixels have cast votes.
The card explains that the Hough Transform works by having each edge pixel 'vote' in a 'parameter space,' and the 'points where many of these curves intersect are local maxima—peaks of votes.' Option A describes template matching, which the card states the Hough Transform was created to overcome due to noise and imperfections.
Read the full bite: Hough Transform: Finding Shapes by Voting
Question 11 of 30
What is the main benefit of applying epipolar geometry when matching points across two images for 3D reconstruction?
Show the answer
Answer: a · It simplifies the search for corresponding points from a 2D area to a specific line.
Epipolar geometry's core function is to constrain the search for a matching point in the second image to an epipolar line, reducing a 2D search to a 1D search. The card explicitly states that the geometry 'will produce errors if objects move independently,' making option C incorrect.
Read the full bite: Epipolar Geometry: Finding 3D Points from 2D Images
Question 12 of 30
Which of the following best describes how the stereo correspondence problem enables a machine to perceive depth?
Show the answer
Answer: b · It identifies the horizontal pixel offset between the same point observed in two different camera images.
The core of stereo correspondence is finding the same point in two images and measuring the horizontal pixel shift, called disparity, which is then used to calculate depth. Option D describes a monocular depth cue, while Option A describes a different technology like Lidar.
Read the full bite: The Stereo Correspondence Problem: How Cameras See in 3D
Question 13 of 30
What fundamental principle allows a disparity map to infer depth from two 2D images?
Show the answer
Answer: b · Detecting the horizontal pixel shift of corresponding points between two camera views.
The card explains that disparity is calculated by finding corresponding pixels in two images and measuring the horizontal distance (shift) between them. This horizontal shift is directly proportional to the object's depth. Option C describes how LiDAR works, which is an alternative depth sensing technology, not the principle behind disparity maps.
Read the full bite: Disparity Map: Seeing Depth from Two 2D Images
Question 14 of 30
Which statement best describes a key limitation of a raw point cloud when compared to a 3D model ready for manufacturing or simulation?
Show the answer
Answer: b · It represents an object as a collection of disconnected points, lacking defined surfaces or volume.
The card explicitly states that a raw point cloud 'has no surfaces, only disconnected' points and 'has no concept of "inside" vs. "outside" or how points connect to each other,' making it unsuitable for tasks requiring defined surfaces or volume. Other options contradict information in the card, such as its ability to store color or its use as a starting point for CAD models.
Read the full bite: Point Cloud: A 3D Shape as a Dust Cloud of Data
Question 15 of 30
When would a convolutional layer be an unsuitable choice for processing data?
Show the answer
Answer: d · Processing customer information in a spreadsheet
The card states convolutional layers are unsuitable for tabular data like customer spreadsheets because local patterns are irrelevant. Image, audio, and video data, however, have inherent spatial or temporal locality, making them ideal.
Read the full bite: Convolutional Layers: Finding Features Anywhere
Question 16 of 30
In a Convolutional Neural Network, what does a single feature map primarily represent?
Show the answer
Answer: b · The presence and location of a specific learned pattern, such as an edge or a texture.
A feature map highlights where a specific pattern (like an edge or curve) appears, showing 'activations' where its specific pattern was found. It does not represent an entire object, which is a common misconception.
Read the full bite: Feature Maps: What a Neural Network 'Sees'
Question 17 of 30
What is the main reason for incorporating a pooling layer into a convolutional neural network?
Show the answer
Answer: b · To decrease the computational load and improve robustness to minor object position changes.
Pooling layers are used to reduce the spatial size of feature maps, which decreases computational cost and memory usage, and they also provide translation invariance, making the model robust to small shifts. Option D is incorrect because pooling reduces spatial resolution and loses fine detail, rather than enhancing it.
Read the full bite: Pooling Layers: Summarizing What Matters in an Image
Question 18 of 30
For a self-driving car to identify every pedestrian and vehicle in its environment, which computer vision task is most appropriate?
Show the answer
Answer: b · Object detection
Object detection is specifically designed to find and box every instance of a specific class in an image, which is crucial for a self-driving car needing to identify all pedestrians and vehicles. Object localization, while finding an object's location, only identifies a single, main object, making it unsuitable for scenarios where multiple instances are critical.
Read the full bite: Localization vs. Detection: One Box or Many?
Question 19 of 30
In object detection, what does an Intersection over Union (IoU) score primarily evaluate?
Show the answer
Answer: b · The degree of spatial overlap between two bounding boxes.
IoU is designed to quantify how well a predicted bounding box spatially aligns with a ground truth box. It does not measure the correctness of the classification label, which is a common misconception, as stated in the card: 'IoU measures spatial overlap, not classification correctness.'
Read the full bite: Intersection over Union (IoU): How Good is Your Bounding Box?
Question 20 of 30
A security system uses frame differencing to detect intruders. Which scenario would most likely cause it to trigger a false alarm?
Show the answer
Answer: d · The camera experiencing slight, continuous vibrations from nearby traffic.
Frame differencing is highly susceptible to camera jitter or shake, which it misinterprets as motion across the entire scene, leading to false positives. While very slow-moving objects are a limitation, they typically result in missed detections (false negatives) rather than false alarms.
Read the full bite: Frame Differencing: The Simplest Way to See Motion
Question 21 of 30
Which situation would most likely cause background subtraction to perform poorly?
Show the answer
Answer: c · A camera mounted on a moving vehicle detecting pedestrians.
The card explicitly states that background subtraction is the "wrong tool when the camera is moving, as there is no stable background to model." A camera mounted on a moving vehicle fits this description. The other options describe scenarios with fixed cameras, which are ideal for this technique.
Read the full bite: Background Subtraction: Finding What's Moving in Video
Question 22 of 30
Optical flow algorithms primarily rely on the assumption of "brightness constancy." Which scenario directly violates this core assumption, leading to inaccurate motion detection?
Show the answer
Answer: a · A sudden change in illumination, such as a shadow passing over an object.
The core assumption of optical flow is that a point's brightness remains constant between frames. A sudden change in illumination, like a shadow, directly violates this by altering a pixel's brightness without actual object motion. While textureless surfaces (Option D) also cause issues, they do not violate brightness constancy itself; rather, they lack distinct patterns for the algorithm to track effectively.
Read the full bite: Optical Flow: Tracking Motion by Watching Pixels
Question 23 of 30
Which task is the most appropriate application for Action Recognition?
Show the answer
Answer: d · Determining if a person in a video is performing a "fall" versus merely "sitting down."
Action Recognition is designed to understand 'what is happening' by analyzing motion over time. Distinguishing between a 'fall' and 'sitting down' requires interpreting a sequence of movements, which is a core application. Counting unique faces or identifying objects/logos in static images or across frames primarily uses object detection or recognition, not the temporal understanding central to action recognition.
Read the full bite: Action Recognition: Teaching Machines to Understand 'Doing'
Question 24 of 30
What is a significant trade-off when incorporating attention mechanisms into a vision model?
Show the answer
Answer: a · It introduces additional computational cost and model complexity.
While attention mechanisms improve focus and accuracy, the card explicitly states they add computational cost and complexity, making the model potentially harder to train and debug. They do not increase training speed, nor do they hinder long-range dependency processing; in fact, self-attention helps with it.
Read the full bite: Attention in Vision: Teaching Models Where to Look
Question 25 of 30
What is the main benefit derived from the adversarial training process in Generative Adversarial Networks (GANs)?
Show the answer
Answer: a · It compels the Generator to create highly realistic and novel data outputs.
The adversarial competition forces the Generator to continuously improve its ability to create new, realistic data to fool the Discriminator. Option B is incorrect because the goal is to generate novel, realistic data, not perfect replicas of existing samples.
Read the full bite: Generative Adversarial Networks (GANs): A Forger and a Detective
Question 26 of 30
Which of the following describes a key limitation of patch embedding in Vision Transformers?
Show the answer
Answer: d · It inherently discards fine-grained spatial details and relationships within each individual patch.
The card explicitly states that patch embedding "discards the fine-grained detail inside each patch" and "loses the precise spatial relationships within each patch." Option A is incorrect because patch embedding's purpose is to make the input sequence manageable, not excessively long.
Read the full bite: Patch Embedding: Turning Images into Words for Transformers
Question 27 of 30
What is the primary role of the "variational" aspect in a Variational Autoencoder (VAE)?
Show the answer
Answer: a · It forces the latent space to be smooth and continuous, enabling meaningful interpolation.
The card explicitly states, "The 'variational' aspect ensures this space is smooth, so you can pick a point near 'cat' and get another valid-looking cat, not garbage." This smoothness is crucial for generating diverse and plausible new data through interpolation. Option C is incorrect because VAEs are known for producing blurrier outputs compared to models like GANs.
Read the full bite: Variational Autoencoders: Learning to Generate Data
Question 28 of 30
Which statement accurately describes the nature of dimensions within a latent space?
Show the answer
Answer: a · They are abstract, model-learned coordinates that lack direct human interpretability.
The card states that latent space dimensions are 'learned by the model, not designed by a person' and 'don't correspond to clean concepts,' making them abstract and not human-interpretable. Option C describes traditional feature engineering, which is distinct from how latent spaces operate.
Read the full bite: Latent Space: A Map Where Distance is Similarity
Question 29 of 30
What is a fundamental difference between a text-to-image model and a traditional image search engine?
Show the answer
Answer: b · Text-to-image models synthesize novel images from descriptions, while search engines locate pre-existing images.
The card explicitly states that a text-to-image model is 'not a search engine for existing images' but rather a 'synthesizer that... generates a brand new image from scratch.' Image search engines, conversely, retrieve existing visual content. Option C is a distractor because while text-to-image models are used for art, they also serve other purposes like marketing and prototyping, and the core difference lies in their output mechanism, not just their primary use case.
Read the full bite: Text-to-Image Generation: From Words to Pixels
Question 30 of 30
When is image captioning the most appropriate AI technique compared to object detection?
Show the answer
Answer: b · To generate a human-readable sentence describing the overall scene and actions.
Image captioning's primary goal is to generate a descriptive, human-readable sentence about an image's content, translating pixels into words. Object detection, while related, is specifically for identifying and locating individual objects, not for generating a narrative description.
Read the full bite: Image Captioning: Teaching Machines to Describe What They See
Could you explain these out loud?
That is what an interview actually tests. Tezvyn gives you questions like these with what the interviewer is really checking, the answer that lands, and the mistake that ends the conversation, in the four minutes before your next meeting.
The iPhone app is on the way
We are building it. Until it lands, nothing here is held back from you: every interview card, your saved cards, streaks and the job board all work in Safari, plus hundreds of free practice quizzes of thirty questions each. Sign in and it all carries over to the app the day it arrives.
Want it as an icon? Tap Share at the bottom of Safari, then Add to Home Screen. It opens full screen and the cards you have read stay available offline.