Summary of Lotus: Learning-based Online Thermal and Latency Variation Management For Two-stage Detectors on Edge Devices, by Yifan Gong et al.
Lotus: learning-based online thermal and latency variation management for two-stage detectors on edge devices
by Yifan Gong, Yushu Wu, Zheng Zhan, Pu Zhao, Liangkai Liu, Chao Wu, Xulong Tang, Yanzhi Wang
First submitted to arxiv on: 1 Oct 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 proposed Lotus framework is a novel approach to dynamically scale CPU and GPU frequencies jointly in an online manner using deep reinforcement learning (DRL) for two-stage object detectors. This is done to avoid thermal throttling and provide stable inference speed on edge devices. The framework is designed specifically for two-stage detectors, which are commonly used in various edge applications, but have high computation costs that can cause severe thermal issues. By jointly scaling CPU and GPU frequencies based on DRL, Lotus aims to reduce latency variation, achieve faster inference, and maintain lower CPU and GPU temperatures under various settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Lotus framework is designed for use on edge devices such as NVIDIA Jetson Orin Nano and Mi 11 Lite mobile platforms. It can help avoid thermal throttling and provide stable inference speed by dynamically scaling CPU and GPU frequencies jointly in an online manner using deep reinforcement learning (DRL). |
Keywords
» Artificial intelligence » Inference » Reinforcement learning
