Главная
Study mode:
on
1
tinyML. Talks Enabling ultra-low Power Machine Learning at the Edge "Once-for-All: Train One Network and Specialize it for Efficient Deployment"
2
Our 1st generation solution
3
New Challenge: Efficient Inference on Diverse Hardware Platfo
4
Our new solution, OFA: Decouple Training and Search
5
Challenge: Efficient Inference on Diverse Hardware Platforms
6
Once-for-All Network: Decouple Model Training and Architecture Design
7
Solution: Progressive Shrinking
8
Connection to Network Pruning
9
Performances of Sub-networks on ImageNe
10
How about search?
11
Accuracy & Latency Improvement
12
More accurate than training from scratch
13
OFA: 80% Top-1 Accuracy on ImageNet
14
Specialized Architecture for Different Hardware Platform
15
AutoML Outperforms Human Designing better MLM 1st place in CVPR 19 Visual Wake Words Challenge
16
What if we also optimize the compiler and run
17
How to save CO2 emission
18
OFA for FPGA Accelerators
19
Next tiny ML Talk
Description:
Explore a groundbreaking approach to efficient machine learning deployment across diverse hardware platforms in this tinyML Talks webcast. Learn about the Once-for-All (OFA) network, which decouples training and search to support various architectural settings. Discover the novel progressive shrinking algorithm, a generalized pruning method that reduces model size across multiple dimensions. Understand how OFA outperforms state-of-the-art NAS methods on edge devices, achieving significant improvements in ImageNet top1 accuracy and latency compared to MobileNetV3 and EfficientNet. Gain insights into OFA's success in the 4th Low Power Computer Vision Challenge and its potential to revolutionize efficient inference on diverse hardware platforms while reducing GPU hours and CO2 emissions.

Train One Network and Specialize It for Efficient Deployment

tinyML
Add to list
#Computer Science #Machine Learning #AutoML #Programming #Cloud Computing #Edge Computing #Model Optimization
1 of 19

tinyML. Talks Enabling ultra-low Power Machine Learning at the Edge "Once-for-All: Train One Network and Specialize it for Efficient Deployment"