Page 274 - Ai Book - 10

7. In neural networks, biases are initialized to large random numbers to ensure a diverse
range of values during weight updates.

8. CNNs take input as one-dimensional arrays and focus primarily on feature extraction rather
than directly operating on images.
D. Very short answer type question.

1. What process is commonly used in Computer Vision to create visual effects by changing pixel values in
an image?
2. What does the term “pixel” represent in the context of digital images?
3. Which technology uses Computer Vision to overlay and embed virtual objects on real-world imagery?

4. What is the primary goal of medical image analysis in the healthcare industry?
5. What does OCR help in extracting from digitized documents and PDFs in digital documentation?
6. In a convolution operation, what is the role of the Kernel matrix in relation to the input image?
7. What does the term “resolution” refer to in the context of digital images?

8. What is the primary difference between Convolutional Neural Networks (CNNs) and ordinary neural
networks in their approach to images?

E. Short answer type question.
1. Explain the concept of “pixel” in digital images and its role in storing visual information.
2. How is the resolution of a digital image defined, and why does it impact image quality?
3. What is the significance of Object Segmentation in computer vision, and how does it differ from other
tasks?

4. Briefly explain the concept of a “Kernel” in the context of convolution and its role in image processing.
5. How does resizing an image impact pixel information, and why is it important in AI systems?
6. Explain the process of saving an image using the imwrite () function in Jupyter Notebook.

7. What is the role of biases in neural networks, and how are they initialized during network setup?
8. In Convolutional Neural Networks (CNNs), how does the convolution operation contribute to feature
extraction in images?

F. Long answer type question.
1. Describe the fundamental concepts of images, including pixels, resolution, and pixel value.
2. How does the convolution process work in Computer Vision, and what is the significance of the Kernel
matrix?

3. Discuss the practical applications of Object Detection in the field of Computer Vision and its relevance in
real-world scenarios.

4. Explain the importance of resizing images in image processing and its implications on pixel information.
5. Explore the role of biases in neural networks, their initialization, and why they are crucial in network
setup.

G. Application based questions.
1. How is Computer Vision utilized in the agricultural sector, specifically in the context of drones?
2. Explore the role of Computer Vision in augmented and mixed reality, emphasizing its impact on computing
devices.

148
148