(A) CNN and ANN:

CNN – Convolutional Neural Network: CNN is a Deep Learning algorithm that takes an input image, assigns importance (learnable weights and biases) to various aspects/objects in the image, and differentiates one from the other. It consists of layers: Convolution Layer, ReLU, Pooling Layer, and Fully Connected Layer.

ANN – Artificial Neural Network: ANN is a computational model inspired by the human brain, consisting of interconnected nodes (neurons) organised in layers that process information to learn patterns and make decisions.

(B) Identification of Layers:

• Box 1 → Hidden Layer (the intermediate processing layer between input and output)
• Box 2 → Output Layer (produces the final prediction/result)

Source: Chapter 5, Section 5.6 Convolutional Neural Networks (CNN)

---

• For (A), the examiner expects the full form + a clear definition for each term. CNN's definition is directly from the textbook (Section 5.6); ANN is a prerequisite concept — keep it concise but accurate.
• For (B), in a standard feedforward neural network diagram, the layers in order are: Input Layer → Hidden Layer(s) → Output Layer. Since Box 1 and Box 2 are the unlabelled middle and last layers respectively, they correspond to Hidden Layer and Output Layer. Always name them precisely — partial credit may be lost for vague answers.
• Do not expand on CNN layers in sub-question (A) beyond what is necessary; save word count as this is only 3 marks.

Computers store RGB images using three separate channels — Red (R), Green (G), and Blue (B). Each channel is a 2D array of pixels, with each pixel value ranging from 0 to 255. All three channels are combined to form the final colour image. Thus, each pixel in an RGB image has three values that together determine its colour.

Source: Chapter 5, RGB Images section

• The examiner expects you to mention: three channels (R, G, B), pixel values 0–255, and that combining all three gives the final colour.
• Avoid writing about grayscale here — stay focused on RGB storage.
• A diagram description (channels stacked) earns full marks if time allows, but the written explanation above is sufficient for 2 marks.

(C) Computer Vision is a superset of Image Processing.

The textbook table explicitly states: "Computer Vision is a superset of Image Processing" and "Image Processing is a subset of Computer Vision." Image Processing enhances/prepares images; Computer Vision extracts meaningful information from them. Option (C) is the only correct differentiator.

(D) The number of pixels in an image is known as resolution.

Resolution refers to the total number of pixels in an image (e.g., expressed as 1280×1024). More pixels = higher resolution = better image quality. Options A and B are factually incorrect; Option C is wrong as resolution depends on pixel count, not file size.

(B) 0 to 255

In the byte image format, pixel values range from 0 to 255, where each pixel is stored as a single byte (8 bits).

Source: Chapter 5, Section 5.5 Convolution

---

The passage explicitly states: "pixels have values varying from 0 to 255 and the value of the pixel determines the color of that pixel." A byte (8 bits) can store 2⁸ = 256 values, i.e., 0 to 255. Option (B) is directly confirmed by the source. Remember: 0 = darkest (black), 255 = brightest (white/full color).

(B) Classification

Classification is the task of assigning an input image one label from a fixed set of categories, which is exactly what identifying and labelling a type of fruit involves.

• Classification assigns a single label to the whole image (e.g., "Mango", "Apple").
• Classification + Localization also finds where the object is — not needed here.
• Object Detection handles multiple objects with locations.
• Segmentation labels individual pixels.
• Since the task only asks to identify the fruit type and assign a label, Classification is correct.

(a) Single objects:
The tasks used for single objects are Classification (assigning one label to the image) and Classification + Localisation (identifying what object is present and where it is located in the image).

(b) Multiple objects:
The tasks used for multiple objects are Object Detection (finding instances of real-world objects in images/videos) and Instance Segmentation (detecting object instances, assigning categories, and labelling each pixel).

Source: Chapter 5, Section 5.2 Computer Vision Tasks

---

• The examiner expects you to correctly pair the task with single/multiple objects.
• Classification + Localisation is explicitly stated in the passage as being "used only for single objects."
• Object Detection and Instance Segmentation deal with multiple object instances.
• Name the task clearly and give a brief one-line description — that earns full marks for a 2-mark question.

(D) Assertion (A) is false, but Reason (R) is true.

The Assertion is false — more pixels means better image quality, as the image more closely resembles the original. The Reason is true — resolution refers to the number of pixels, determining clarity and detail.

The textbook clearly states: "The more pixels you have, the more closely the image resembles the original." So the Assertion is factually incorrect. The Reason correctly defines resolution. Since A is false and R is true, option (D) is correct. Always check Assertion and Reason independently before picking an option.

(B) Computer Vision

The passage under 5.1 Applications of Computer Vision explicitly states: "Modern-day apps like Instagram and Snapchat have a lot of features based on the usage of computer vision. The application of face filters is one among them." So (B) is the only correct option.

(D) Clustering

Clustering is an unsupervised learning technique where data points are grouped based on similarity without predefined labels. The diagram showing ungrouped data points being divided into distinct groups represents clustering.

(D) Classification + Localisation

This task involves both identifying what object is present in an image and determining its precise location within the image.

Source: Chapter 5, Section 5.2 Computer Vision Tasks

The passage explicitly states: "Classification + Localisation is the task that involves both processes of identifying what object is present in the image and at the same time identifying at what location that object is present." Object Detection (B) also finds locations, but for multiple objects; Classification + Localisation is specifically defined as doing both identification and location for a single object — matching the question exactly.

Two key impacts of Computer Vision on medical imaging are:

1. Conversion of 2D scans to 3D models: It reads and converts 2D scan images into interactive 3D models, enabling medical professionals to gain a detailed understanding of a patient's health condition.

1. Assisting doctors in interpretation: It acts as an assistant to physicians by not only creating and analysing medical images but also helping doctors interpret them accurately.

Source: Applications of Computer Vision, Chapter 5

---

• The question is 2 marks, so two distinct points are needed — one per mark.
• Both points must come directly from the "Medical Imaging" paragraph in Section 5.1.
• Examiners look for: (a) 3D model conversion from 2D scans, and (b) the role as a diagnostic assistant/interpreter for doctors.
• Avoid vague answers like "it helps in hospitals" — be specific to what the textbook states.

| Basis | Computer Vision (CV) | Natural Language Processing (NLP) |
|-------|---------------------|----------------------------------|
| Definition | Enables machines to interpret and understand visual information (images/videos). | Enables machines to understand and process human language (text/speech). |
| Input | Images, videos | Text, audio/speech |
| Example | Facial recognition, object detection | Chatbots, language translation |

Source: Unit 5: Computer Vision, Chapter 5

• Examiners expect a tabular format for "differentiate" questions — it's cleaner and earns full marks faster.
• You need at least 2 clear points of difference for 2 marks.
• CV focuses on visual data; NLP focuses on language data — this is the core distinction to highlight.
• Even though NLP is not covered in the source passages, CBSE expects you to know this distinction from general AI knowledge introduced earlier in the course.

(C) Statement 1 is correct but Statement 2 is incorrect.

Statement 1 is correct. Statement 2 is incorrect because image-based search finds items, people, and places by giving their images (not sounds) to the system.

The key error in Statement 2 is the word "sounds" — image-based search uses visual/image input, not audio. Statement 1 is well-supported by the passage (Google's Search by Image uses Computer Vision). Examiners expect you to identify the exact factual error.

(B) Classification + Localization

Classification + Localization involves identifying what object is present and its location, and is used only for single objects.

Source: Chapter 5, Section 5.2 Computer Vision Tasks

The passage explicitly states: "It is used only for single objects" for Classification + Localization. Object Detection handles multiple objects, and Instance Segmentation deals with pixel-level labeling of multiple instances. "Non-Localization" is not a real CV task. Examiners expect the correct option with a brief justification from the text.

(B) Extracting meaningful information from images

Image processing focuses on processing raw input images to enhance them or prepare them for further tasks, while extracting meaningful information is the broader goal of Computer Vision.

The question asks specifically about "image processing" in the context of Computer Vision. According to the source, image processing is a subset of Computer Vision focused on processing/enhancing raw images — but the correct option (B) reflects the overall purpose of Computer Vision tasks that image processing supports. Examiners expect students to pick (B) as it best matches the described role. Note the distinction: image processing prepares images; Computer Vision extracts meaning.

Source: Chapter 5, Section 5.1 (Computer Vision Vs. Image Processing table)

| Feature | Grayscale Image | RGB Image |
|---|---|---|
| Channels | Single channel (1 plane) | Three channels — Red, Green, Blue |
| Pixel value | One value per pixel (0–255) | Three values per pixel (R, G, B) |
| Storage size | Height × Width | Height × Width × 3 |
| Representation | Shades of gray (black to white) | Full colour image |

A grayscale image stores each pixel using 1 byte (single intensity value), while an RGB image stores each pixel using 3 values combining red, green, and blue intensities to produce colour.

Source: Chapter 5, Sections — Grayscale Images & RGB Images

---

• Examiners expect a comparison format (table or two clear points) for "differentiate" questions.
• Key contrast points: number of channels, values per pixel, and how colour is represented.
• Remember: grayscale pixel size = 1 byte; RGB pixel = 3 bytes (one per channel). This is a commonly tested fact.
• Avoid writing long paragraphs — a neat table with 3–4 rows earns full marks efficiently.

(b) Computer Vision

The passage explicitly states that face filters on apps like Instagram and Snapchat use Computer Vision to identify facial dynamics and apply the selected filter. This is a direct fact from Section 5.1 Applications of Computer Vision.

(a) Both (A) and (R) are true and (R) is the correct explanation for (A).

The term "resolution" refers to the number of pixels in an image, expressed as width × height (e.g., 1280×1024), so (R) correctly explains (A).

Source: Chapter 5, Section 5.2 – Resolution

---

The textbook explicitly states: "The number of pixels in an image is sometimes called the resolution… one convention is to express resolution as the width by the height, for example, 1280×1024." Both the Assertion and Reason are accurate, and the Reason directly defines/explains the Assertion — so option (a) is correct. Note: the Reason says "height × width" while the book says "width × height" — both express the same idea and are acceptable.

(c) Instance Segmentation

Instance Segmentation detects multiple objects, assigns each a category, and labels every pixel — making it suitable for handling multiple objects simultaneously.

Source: Computer Vision Tasks, Chapter 5

The passage states Classification + Localisation is used only for single objects, while Instance Segmentation handles detecting instances (multiple objects) and labels each pixel. Object Detection also handles multiple objects, but it is not among the options; Instance Segmentation is the best answer from the given choices.

Face lock (facial recognition) is a computer vision application where the smartphone's camera captures the user's face and converts it into digital data. The algorithm analyses facial features (like eyes, nose, jawline) and compares them with the stored facial data. If it matches, the phone unlocks; otherwise, access is denied.

Source: Applications of Computer Vision, Chapter 5

---

• The question asks you to briefly discuss, so 2–3 sentences covering what it does and how it works is sufficient.
• Key points examiners expect: camera captures face → features analysed → compared with stored data → match/no match decision.
• Avoid writing a long paragraph; 40–55 words is ideal for 2 marks.
• The textbook specifically mentions facial recognition under security applications — link your answer to that context.

CV stands for Computer Vision — a branch of Artificial Intelligence that enables machines to interpret and understand visual information from the real world.

The question asks for the full form of CV in the AI domain. From the textbook, Unit 5 is titled Computer Vision and the term CV is used in section 5.2 in reference to image classification tasks. Examiners expect the full form plus a one-line definition for context in a 1-mark answer.