Summary of Ondev-lct: On-device Lightweight Convolutional Transformers Towards Federated Learning, by Chu Myaet Thwal et al.
OnDev-LCT: On-Device Lightweight Convolutional Transformers towards federated learning
by Chu Myaet Thwal, Minh N.H. Nguyen, Ye Lin Tun, Seong Tae Kim, My T. Thai, Choong Seon Hong
First submitted to arxiv on: 22 Jan 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes a new type of model called OnDev-LCT, which is designed to be lightweight and efficient for on-device vision tasks. This is achieved through the use of image-specific inductive biases and efficient depthwise separable convolutions. The model is tested on benchmark image datasets and outperforms existing lightweight vision models while having fewer parameters and lower computational demands. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re trying to train a machine learning model on lots of different devices, like smartphones or computers. This is called federated learning (FL). To make it work, you need models that are good at handling the unique challenges of distributed learning. Some models, like Vision Transformers (ViT), are really good but they’re too big and use too much computer power to work on most devices. The researchers in this paper created a new model called OnDev-LCT that’s designed specifically for these kinds of devices. It uses special tricks like efficient convolutions and attention mechanisms to be fast, efficient, and effective. |
Keywords
* Artificial intelligence * Attention * Federated learning * Machine learning * Vit
