Summary of Mtl-split: Multi-task Learning For Edge Devices Using Split Computing, by Luigi Capogrosso et al.
MTL-Split: Multi-Task Learning for Edge Devices using Split Computing
by Luigi Capogrosso, Enrico Fraccaroli, Samarjit Chakraborty, Franco Fummi, Marco Cristani
First submitted to arxiv on: 8 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); Distributed, Parallel, and Cluster Computing (cs.DC)
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 explores the emerging approach of Split Computing (SC), which involves dividing a Deep Neural Network (DNN) to be deployed on both edge devices and remote servers. This allows for the power of DNNs to be harnessed in latency-sensitive applications where local computation is insufficient, but only parts of the network can be deployed remotely. The paper also focuses on Multi-Task Learning (MTL), where a single DNN handles multiple inference tasks instead of dedicated models for each task. To address this problem, the authors propose MTL-Split, an architecture that shows promising results on synthetic and real-world data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Split Computing allows you to use Deep Neural Networks in devices with limited power. This is important because some applications need answers fast, but can’t do all the calculations themselves. Instead, they can split the job between a device and a remote server. The paper also talks about Multi-Task Learning, where one neural network does many jobs instead of separate networks for each task. The authors came up with a new way to do this called MTL-Split, which works well on fake and real data. |
Keywords
» Artificial intelligence » Inference » Multi task » Neural network
