Activity — Test your knowledge on diffusion models, visual AI, and image generation techniques
Instructions
Answer all questions to complete the activity
No time limit — take your time to think
Mix of MCQ, True/False, Fill-in-the-blank, Match & Timeline
Your score is tracked automatically
Your Progress
0 / 35 answeredScore: 0
Multiple Choice Questions
MCQ
Q1. What do diffusion models do?
Correct! Diffusion models generate images by gradually removing noise from random static to produce coherent visuals.
MCQ
Q2. What is latent space in diffusion models?
Correct! Latent space is a compressed representation of images where the diffusion model performs most of its processing, making computation more efficient.
MCQ
Q3. What does ControlNet enable?
Correct! ControlNet adds conditional control to diffusion models, enabling users to guide generation with structural inputs like edges, depth maps, or pose skeletons.
MCQ
Q4. What are negative prompts?
Correct! Negative prompts specify elements you want to exclude from the generated image, helping refine outputs.
MCQ
Q5. What is inpainting?
Correct! Inpainting is the process of filling in selected or missing regions of an image using AI to generate contextually appropriate content.
MCQ
Q6. What is outpainting?
Correct! Outpainting extends an image beyond its original boundaries, generating new content that seamlessly continues the existing scene.
MCQ
Q7. What is img2img?
Correct! img2img uses an existing reference image as the starting point for generation, allowing style transfer and modifications while preserving composition.
MCQ
Q8. Which company developed DALL-E?
Correct! DALL-E was developed by OpenAI. The name is a blend of Salvador Dalí and WALL-E.
MCQ
Q9. What is the Subject-View-Style framework for?
Correct! The Subject-View-Style framework helps structure prompts by defining the subject, camera view, and artistic style for consistent, high-quality image generation.
MCQ
Q10. What is the first step in the diffusion process?
Correct! The diffusion process begins with pure random noise, which is then progressively denoised into a coherent image.
MCQ
Q11. What does "denoising" mean in diffusion models?
Correct! Denoising is the progressive removal of noise from a random signal, gradually revealing a clear, coherent image.
MCQ
Q12. Which is open-source for local image generation?
Correct! Stable Diffusion by Stability AI is open-source, allowing users to run image generation locally on their own hardware.
MCQ
Q13. What is text-to-video generation?
Correct! Text-to-video generation creates animated video sequences from textual descriptions, extending image generation into the temporal dimension.
MCQ
Q14. What is prompt rewriting?
Correct! Prompt rewriting involves refining and optimizing prompts to improve the quality and accuracy of AI-generated outputs.
MCQ
Q15. Which principle involves adding reference images to prompts?
Correct! The "Provide Examples" principle includes adding reference images (img2img) to guide the model toward the desired output style or composition.
True / False
True / False
Q16. Diffusion models work by adding noise to images.
Correct! False. Diffusion models work by removing noise. The training process adds noise, but generation reverses it by progressively denoising.
True / False
Q17. Negative prompts help exclude unwanted elements.
Correct! True. Negative prompts specify what you do NOT want in the generated image, helping refine and clean up outputs.
True / False
Q18. Inpainting can only fill the center of images.
Correct! False. Inpainting can fill any selected region of an image, not just the center. You can mask and regenerate any area.
True / False
Q19. Stable Diffusion is a proprietary model.
Correct! False. Stable Diffusion is open-source, allowing anyone to download, modify, and run it locally.
True / False
Q20. ControlNet enables structural control over generation.
Correct! True. ControlNet allows structural inputs like edge maps, depth maps, and pose skeletons to guide the generation process.
True / False
Q21. Outpainting extends images beyond their original borders.
Correct! True. Outpainting generates new content that extends an image beyond its original boundaries while maintaining consistency.
True / False
Q22. img2img requires no reference image.
Correct! False. img2img fundamentally requires a reference image as the starting point for generation.
True / False
Q23. Temperature affects randomness in image generation.
Correct! True. Temperature controls the randomness of the generation process — higher values produce more creative/diverse outputs, lower values are more deterministic.
Fill in the Blank
Fill in the Blank
Q24. ______ models generate images by reversing a noise-adding process.
Correct! Diffusion models generate images by progressively removing noise from random static.
Fill in the Blank
Q25. ______ prompts tell the model what elements to exclude.
Correct! Negative prompts specify elements to exclude from the generated image.
Fill in the Blank
Q26. ______ fills in missing or selected parts of an image.
Correct! Inpainting is the technique of filling in masked or missing regions of an image.
Fill in the Blank
Q27. ______ extends an image beyond its original boundaries.
Correct! Outpainting generates new content that extends an image beyond its original borders.