Summary of Deepracer on Physical Track: Parameters Exploration and Performance Evaluation, by Sinan Koparan et al.
DeepRacer on Physical Track: Parameters Exploration and Performance Evaluation
by Sinan Koparan, Bahman Javadi
First submitted to arxiv on: 6 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Robotics (cs.RO)
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 explores the capabilities of AWS DeepRacer’s physical racetrack, examining the impact of hyperparameters on performance and object avoidance. Two experiments were conducted: Experiment I evaluated the effects of different hyperparameters, including gradient descent batch size, loss type, and training time settings, while Experiment II focused on object avoidance in the physical environment. The results show that models with higher gradient descent batch sizes performed better in simulations but preferred a batch size of 128 in the physical environment. Models using Huber loss outperformed those using MSE loss in both environments. However, object avoidance remained challenging when bringing simulated models to the real world. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary AWS DeepRacer’s physical racetrack was tested to see how different settings affect its performance. The researchers changed things like how fast it learns and what kind of mistakes it makes. They also tried to get the robot to avoid objects while racing. The results showed that some settings made it better in simulations, but not as good in real life. It’s still hard for the robot to avoid objects in the physical environment. |
Keywords
» Artificial intelligence » Gradient descent » Mse
