Question 224 MCQ · 5.0 marks
(a) and (c)
(b) and (d)
(a), (c), and (d)
All of these
A published solution is not available for this question yet.
Deep Learning Practice · End Term · Sep 2024 · AN
Deep Learning Practice · End Term · Sep 2024 AN
← Course papers · Start practice / exam
Questions and published explanations below are available without starting a test. Some questions may not have a published solution yet.
0 / 20 published solutions
Question 225 MCQ · 5.0 marks
LeNet
ResNet
AlexNet
VGG16
None
A published solution is not available for this question yet.
Question 226 MCQ · 5.0 marks
Suppose I have an image of size 227×227. Which of the following code snippets correctly
implements a Min Pooling operation with a window size of 2 × 2, stride of 2, and padding of 1 in
PyTorch?
A published solution is not available for this question yet.
Question 227 MCQ · 5.0 marks
Consider a black-and-white image of dimension 227×227. Which of the following correctly
demonstrates a function to flatten it along with the dimension of the newly formed vector?
A published solution is not available for this question yet.
Question 228 MCQ · 5.0 marks
Which of the following code snippets correctly implements the skip connection in ResNet?
A published solution is not available for this question yet.
Question 229 MCQ · 5.0 marks
11
10
12
14
A published solution is not available for this question yet.
Question 230 MCQ · 5.0 marks
B, A, C, D
A, B, C, D
B, C, D, A
C, A, B, D
A published solution is not available for this question yet.
Question 231 MCQ · 5.0 marks
Which of the following code snippets correctly models the loss function in YOLO?
A published solution is not available for this question yet.
Question 232 MCQ · 5.0 marks
Consider the following predictions and ground truths for a binary classification problem:
- True Positives (TP): 30 - False Positives (FP): 10 - False Negatives (FN): 20
Which of the following correctly calculates the Precision and F1-Score for this classification task?
A published solution is not available for this question yet.
Question 233 MCQ · 5.0 marks
How many convolutional layers are there in the original YOLO architecture?
24
19
53
75
A published solution is not available for this question yet.
Question 234 MCQ · 5.0 marks
0.142
0.250
0.333
0.500
A published solution is not available for this question yet.
Question 235 MCQ · 5.0 marks
Block A generates high-level global features and outputs an initial coarse
depth map, while Block B refines the depth map using local details from the RGB image and the
coarse depth map.
Block A performs refinement of the depth map using concatenated coarse
depth and RGB features, while Block B generates the coarse depth map from the input RGB
image.
Both Block A and Block B are coarse networks, with Block B performing an
additional refinement step.
Block A generates a low-resolution coarse depth map but does not include
global features, while Block B produces fine-grained depth without refinement.
A published solution is not available for this question yet.
Question 236 MCQ · 5.0 marks
Which of the following code snippets correctly implements the **downsampling** operation in a
U-Net architecture?
A published solution is not available for this question yet.
Question 237 MCQ · 5.0 marks
Consider the following predictions and ground truth values:
- Predictions: [3.0,−0.5, 2.0, 7.0] - Ground Truth: [2.5, 0.0, 2.0, 8.0]
The Mean Squared Error (MSE) is calculated using one of the following code snippets. Identify the
correct implementation and its result.
A published solution is not available for this question yet.
Question 238 MCQ · 5.0 marks
In a stereo vision system, left-right disparity is used to compute depth by comparing
corresponding points in the left and right images. Which of the following statements about left
right disparity is correct?
Disparity is the absolute difference in pixel intensities between the left and
right images.
Disparity increases as the depth of an object increases from the cameras.
Disparity is the horizontal shift between corresponding points in the left and
right images.
Disparity is the vertical difference between points in the left and right images.
A published solution is not available for this question yet.
Question 239 MCQ · 5.0 marks
What are the primary roles of the **generator** and **discriminator** in a Super-Resolution GAN
(SRGAN) architecture?
The generator produces high-resolution images from low-resolution inputs,
while the discriminator distinguishes real high-resolution images from generated high-resolution
images.
The generator distinguishes real high-resolution images from generated
images, while the discriminator produces high-resolution images from low-resolution inputs.
Both the generator and discriminator produce high-resolution images from
low-resolution inputs.
Both the generator and discriminator distinguish real high-resolution images
from generated high-resolution images.
A published solution is not available for this question yet.
Question 240 MCQ · 5.0 marks
Which of the following code snippets correctly adds gaussian noise to an image?
A published solution is not available for this question yet.
Question 241 MCQ · 5.0 marks
What are the benefits of using depthwise and pointwise convolutions (as in depthwise separable
convolutions) compared to standard convolutions?
They reduce the number of parameters and computational cost, making the
model more efficient.
They improve the accuracy of the model by learning richer spatial features.
They reduce overfitting by performing regularization during convolution
operations.
They allow convolutions to operate across both spatial dimensions and depth
channels simultaneously.
A published solution is not available for this question yet.
Question 242 MCQ · 5.0 marks
Which of the following code snippets correctly implements the **generator** for a super-
resolution GAN (SRGAN) network?
A published solution is not available for this question yet.
Question 243 MSQ · 5.0 marks
Which of the following are common use cases of super-resolution using SRGAN?
Enhancing medical imaging scans for improved diagnosis.
Increasing the resolution of satellite images for geographical analysis.
Generating high-quality images for video streaming and upscaling.
Translating text from one language to another in images.
Improving the resolution of historical or archival images.
**SPG**
**Section Id :** 64065375581
**Section Number :** 11
**Section type :** Online
**Mandatory or Optional :** Mandatory
**Number of Questions :** 19
**Number of Questions to be attempted :** 19
**Section Marks :** 35
**Display Number Panel :** Yes
**Section Negative Marks :** 0
**Group All Questions :** No
**Enable Mark as Answered Mark for Review and**
No
**Clear Response :**
**Section Maximum Duration :** 0
**Section Minimum Duration :** 0
**Section Time In :** Minutes
**Maximum Instruction Time :** 0
A published solution is not available for this question yet.