Introduction to On-Device AI - Qualcomm (DeepLearningAI)

Overview

This is a comprehensive set of notes and resources for a crash course on deploying AI models on edge devices, provided by DeepLearning.AI and taught by Krishna Sridhar from Qualcomm.

What you'll Learn

• Learn to deploy AI models on edge devices like smartphones, using their local compute power for faster and more secure inference.
• Explore model conversion by, converting your PyTorch/TensorFlow models for device compatibility, and quantize them to achieve performance gains while reducing model size.
• Learn about device integration, including runtime dependencies, and how GPU, NPU, and CPU compute unit utilization affect performance.

Prerequisites

This course is designed for beginner AI developers, ML engineers, data scientists, and mobile developers looking to deploy optimized models on edge devices. Familiarity with Python, as well as PyTorch or TensorFlow is recommended.

Course Outline

• Introduction
• Why On-Device?
• Deploying Segmentation Models On-Device
• Preparing for On-Device Deployment
• Quantizing Models
• Device Integration
• Conclusion

Lab: Chapters & Notebooks

Chapters	Notebooks	Demos
Introduction	-	-
Why On-Device?	-	-
Deploying Segmentation Models On-Device	L2_Student.ipynb	- ffnet_40 (on-device) - ffnet_54s (on-device) - ffnet_78s (on-device) - ffnet_78s_lowres (on-device) - ffnet_122s_lowres (on-device)
Preparing for On-Device Deployment	L3_Student.ipynb	-
Quantizing Models	L4_Student.ipynb	-
Device Integration (Final App)	App/Project Guidelines	Final Demo
Conclusion	-	-

Disclaimer

For important details about this repository's content, please review the DISCLAIMER.md.

References

• Main course - DeepLearning.AI