Computer Vision Interview
20 essential Q&A
Updated 2026
segmentation
Segmentation Overview: 20 Essential Q&A
Semantic vs instance, classical region methods, and how we evaluate masks.
~11 min read
20 questions
Intermediate
semanticinstancewatershedIoU
Quick Navigation
1
What is image segmentation?
⚡ easy
Answer: Partitioning an image into regions or labeling each pixel with a category—bridges low-level pixels and high-level objects.
2
What is semantic segmentation?
📊 medium
Answer: Each pixel gets a class label (road, sky, person) without distinguishing different instances of the same class.
3
What is instance segmentation?
📊 medium
Answer: Separate individual objects even within same class—each instance has its own mask and ID.
4
What is panoptic segmentation?
🔥 hard
Answer: Unifies stuff (amorphous regions like sky) and things (countable objects) with non-overlapping masks covering the whole image.
5
Is thresholding segmentation?
⚡ easy
Answer: Yes for binary foreground/background—simplest form; limited for complex scenes without additional cues.
6
What is region growing?
📊 medium
Answer: Start from seeds, merge neighboring pixels similar under a criterion (intensity, texture)—sensitive to seed placement and noise.
7
Split and merge?
📊 medium
Answer: Recursively split non-uniform regions, then merge adjacent similar regions—quadtree-style classical approach.
8
Segment with k-means in color space?
📊 medium
Answer: Cluster pixel colors (RGB or LAB); each cluster is a segment—produces patchy results without spatial smoothness unless augmented.
9
What is mean shift segmentation?
🔥 hard
Answer: Mode-seeking in joint color-spatial feature space—clusters pixels to local density peaks; smooths labels but can be slow.
10
What is the watershed transform?
📊 medium
Answer: Treat gradient magnitude as height map; flood from markers—without markers causes oversegmentation; marker-controlled watershed is common.
11
What are graph cuts?
🔥 hard
Answer: Pixels as nodes, pairwise smoothness + unary data costs; find min-cut for globally good binary partition—used in GrabCut-style energy minimization.
12
What is GrabCut?
📊 medium
Answer: Iterative graph-cut segmentation with Gaussian mixture models on RGB—user provides box or strokes; refines foreground/background.
13
What are active contours (snakes)?
📊 medium
Answer: Deform curve minimizing internal smoothness + external edge attraction—classic for medical boundaries; level-set extensions handle topology changes.
14
Oversegmentation?
⚡ easy
Answer: Too many small regions—watershed without markers; fix with merging, markers, or learned superpixels (SLIC).
15
Evaluate masks with IoU?
📊 medium
Answer: Intersection over union per class or instance; mean IoU (mIoU) standard for semantic segmentation benchmarks.
16
Boundary F-score?
🔥 hard
Answer: Measures alignment of predicted vs GT contours—complements IoU for thin structures.
17
Deep learning for segmentation?
📊 medium
Answer: FCN replaces FC layers with convs; U-Net encoder-decoder with skip connections—dominant paradigm now with transformers emerging.
18
Video segmentation?
📊 medium
Answer: Temporal consistency, optical flow warping, or memory networks—object masks tracked across frames (VOS).
19
Interactive segmentation?
⚡ easy
Answer: User clicks/scribbles guide model (GrabCut, deep interactive)—few-shot refinement for editing.
20
Pick classical vs deep?
📊 medium
Answer: Classical: fast, little data, controlled scenes. Deep: cluttered natural images, need labels and compute—often hybrid for industrial + DL refinement.
Segmentation Cheat Sheet
Types
- Semantic
- Instance
- Panoptic
Classical
- Watershed
- Graph cut
- GrabCut
Metrics
- mIoU
- Boundary F
💡 Pro tip: Define semantic vs instance before diving into algorithms.
Full tutorial track
Go deeper with the matching tutorial chapter and code examples.