CNN CNN - Convolutional Neural Networks Semantic Segmentation Questions and Answers

Question 1

What is the goal of semantic segmentation in computer vision?

Accepted Answer

Assigning a class label to every pixel in an image

Answer

Detecting bounding boxes around objects

Answer

Identifying unique instances of each object

Answer

Generating captions for images

Question 2

What key innovation did Fully Convolutional Networks (FCN) introduce for semantic segmentation?

Accepted Answer

Replacing fully connected layers with convolutional layers to enable end-to-end pixel-wise prediction

Answer

Using fully connected layers for dense prediction

Answer

Adding attention mechanisms

Answer

Using recurrent layers for sequential processing

Question 3

What architecture is U-Net primarily designed for, and what is its signature structural feature?

Accepted Answer

Biomedical image segmentation with a symmetric encoder-decoder and skip connections

Answer

Object detection with anchor boxes

Answer

Image classification with residual blocks

Answer

Video segmentation with temporal convolutions

Question 4

What is the role of skip connections in the U-Net architecture?

Accepted Answer

They pass high-resolution feature maps from the encoder to the decoder to recover spatial detail lost during downsampling

Answer

They skip residual computations to speed up training

Answer

They skip low-performing layers during inference

Answer

They connect non-adjacent layers for gradient flow only

Question 5

What is the fundamental difference between semantic segmentation and instance segmentation?

Accepted Answer

Semantic segmentation assigns the same label to all pixels of a class, while instance segmentation distinguishes individual object instances

Answer

Semantic segmentation is faster

Answer

Instance segmentation works only on videos

Answer

Semantic segmentation requires depth data

Question 6

What is the encoder-decoder architecture in CNN-based segmentation networks?

Accepted Answer

A network where the encoder downsamples the input to a bottleneck and the decoder upsamples it back to the original resolution for pixel prediction

Answer

An autoencoder used for image compression

Answer

A two-stage classifier followed by a regressor

Answer

A GAN-based structure for generating masks