CNN Architecture and Operations — past-year questions
Deep Learning (CSC701) · Semester 7 · Module 4 · AIML & Data Science
8 past-year questions on CNN Architecture and Operations have appeared in Mumbai University CSC701 papers between 2023–2026, 9 times in total. Every question is shown in full below with its marks and repeat count. Model answers require a free account.
Questions asked on CNN Architecture and Operations
- 1.
Explain Pooling operation in CNN.
5M2× seenlast asked 2024 - 2.
Explain the complete workflow of a Convolutional Neural Network with a neat diagram and suitable example.
10M1× seenlast asked 2026 - 3.
Explain basic working of CNN.
5M1× seenlast asked 2025 - 4.
Consider a CNN block with the following configuration: Input: 16 channels with spatial size 128x128; Conv-1: 32 filters of size 5x5, stride = 1, padding = 2 (no dilation); MaxPool: 3x3 window, stride = 3, padding = 0; Conv-2: 64 filters of size 3x3, stride = 2, padding = 1 (no dilation). Answer the following: (i) Compute the spatial size (HxW) of the output after Conv-1. (ii) Compute the spatial size (HxW) of the output after the MaxPool layer. (iii) Compute the spatial size (HxW) of the output after Conv-2.
10M1× seenlast asked 2025 - 5.
Explain weight sharing in CNN.
5M1× seenlast asked 2025 - 6.
Consider a CNN layer with the following configuration: -The input to the layer has 32 channels and a spatial size of 64x64. -The convolutional layer has 64 filters (kernels), each of size 3x3, with a stride of 1 and no padding. -Each filter is applied to every channel of the input. Calculate the total number of parameters (weights) in this convolutional layer.
10M1× seenlast asked 2024 - 7.
Explain the architecture of CNN with the help of a diagram.
10M1× seenlast asked 2024 - 8.
Explain CNN architecture in detail. Suppose, we have input volume of 32 × 32 × 3 for a layer in CNN and there are ten 5 × 5 filters with stride 1 and pad 2; calculate the number of parameters in this layer of CNN.
10M1× seenlast asked 2023
Want the answers to these questions?
The questions above are free to read. Create a free account to get AI model answers for Deep Learning, a curated set of the questions that matter most, flashcard revision, and a study plan built around your exam date.