Prompt Engineering for Vision Models (DeepLearning.AI)

(Source: SAM)

Overview

These notes and resources are compiled from the crash course Prompt Engineering for Vision Models like Meta's SAM (Segment Anything Model) or Stable Diffusion, offered by DeepLearning.AI.

The course, led by Andrew Ng and instructors from Comet (Abby Morgan, Jacques Verré, and Caleb Kaiser), explores techniques for prompting vision models like image generation and object detection.

Key Concepts

• Gain a foundational understanding of prompt engineering techniques for guiding vision models.
• Explore methods for image generation, object detection, and image segmentation using text prompts.
• Learn to fine-tune diffusion models for personalized image creation with DreamBooth.
• Discover best practices for experimenting and tracking progress in prompt engineering workflows.

Course Contents

• Lesson0: Introduction
• Lesson1: Overview
• Lesson2: Image Segmentation
• Lesson3: Object Detection
• Lesson4: Image Generation
• Lesson5: Fine-tuning
• Lesson6: Conclusion

Setup & Requirements

Requirements

• All you need is a Deep LearningAI user account to start learning for free.

Lab: Hands-on Exercises

Chapter	Notebook
Lesson0: Introduction	-
Lesson1: Overview	-
Lesson2: Image Segmentation
Lesson3: Object Detection
Lesson4: Image Generation
Lesson5: Fine-tuning

References

Main Course :

• https://learn.deeplearning.ai/courses/prompt-engineering-for-vision-models/lesson/1/introduction

Others short Free Courses available on DeepLearning.AI :

• https://learn.deeplearning.ai/

Resources:

• Awesome-Diffusion-Models: A collection of resources and papers on Diffusion Models