Summary of Actnas : Generating Efficient Yolo Models Using Activation Nas, by Sudhakar Sah et al.
ActNAS : Generating Efficient YOLO Models using Activation NAS
by Sudhakar Sah, Ravish Kumar, Darshan C. Ganji, Ehsan Saboori
First submitted to arxiv on: 11 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Neural and Evolutionary Computing (cs.NE)
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 This paper investigates the effects of using mixed activation functions in YOLO-based models for computer vision tasks. The authors explore different combinations of ReLU, SiLU, SELU, and other functions to optimize performance while minimizing latency and memory usage across various edge devices, including CPUs, NPUs, and GPUs. They propose a novel approach leveraging Neural Architecture Search (NAS) to design optimal mixed-activation models and demonstrate a slight improvement in mean Average Precision (mAP) compared to baseline SiLU models, with significant reductions in processing time and memory consumption. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how different “wake-up” functions help artificial intelligence networks learn better. Right now, most AI networks use the same function throughout the whole system. The researchers wanted to see what would happen if they used different functions for different parts of the network. They tested this idea on a specific type of AI called YOLO (You Only Look Once), which is good at recognizing objects in images. They found that using the right mix of functions could make the AI faster and use less memory, while still being just as accurate. |
Keywords
» Artificial intelligence » Mean average precision » Relu » Yolo
