Completed
Solving the Imbalance with Patch-based Inference
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
TinyML for All: Full-stack Optimization for Diverse Edge AI Platforms
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 TinyML is about Constraints
- 3 Everything Together: Real-world Al on Tiny MCUS
- 4 Brief History of MCUNets
- 5 Opportunity in Fundamental ML Algorithms
- 6 New Problem: Imbalanced Memory Distribution of CNNS
- 7 Solving the Imbalance with Patch-based Inference
- 8 MCUNet-v2 Takeaways
- 9 Once-for-All Network
- 10 Problem in Training for Tiny Models
- 11 NetAug for TinyML
- 12 Problem: Training Memory is much larger
- 13 TinyTL: Up to 6.5x Memory Saving without Accuracy Loss
- 14 Differentiable Augmentation
- 15 TinyML for LIDAR & Point Cloud
- 16 Full Stack LIDAR & Point Cloud Processing
- 17 Takeaways: Coming Back to MCUNets
- 18 Fundamental Problems in TinyML
- 19 OmniML "Compress" the Model Before Training
- 20 OmniML: Enable TinyML for All Vision Tasks
- 21 Founding Team
