sabanaherons2024
Science Score: 26.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (9.1%) to scientific vocabulary
Last synced: 10 months ago
·
JSON representation
Repository
Basic Info
- Host: GitHub
- Owner: SantiagoSantafe
- License: other
- Language: C++
- Default Branch: master
- Size: 1.78 GB
Statistics
- Stars: 1
- Watchers: 2
- Forks: 1
- Open Issues: 0
- Releases: 0
Created over 2 years ago
· Last pushed almost 2 years ago
Metadata Files
Readme
License
Citation
README.md
Sabana-Herons Code Release
This is the 2024 Sabana-herons code release. this code is heavily based in B-Human 2023 code release available in as we are a small team mainly composed by undergrad students.
Nonetheless, we have added our own improvements to the code, and behaviors that adapt the best to our strategy, therefore, the following list is meant to describe those improvements and adaptations we've done:
Improvements:
- 5vs5 tactics and strategies: Due to our number of available NaoV6 robots and team size, we're currently participating exclusively in SPL Shield Challenge, we've adapted and created B-Human code to work in 5vs5 matches.
- Fail analysis with back head button: Due to the difficulty to rapidly and consistently detect fails in our current field setup (a kind of true official game field but - which due to installation can generate some problems in robots motion schema), we've developed a functionality that when can output the most "Critical recent fails" so we don't have to access each robot's log.
- Improved DribbleToGoal Skill: We noted DribbleToGoal skill may encounter some edge cases for Striker, especially when arriving to the delimitation lines of field, which caused the striker to lose the ball and generate a Goal Kick In for adversaries or to shot at the post of the goal, we tried to improve that by almost never letting the striker shot the ball outside field boundary generating a KickIn or striking the Goal Post.
- Optimize the run time of 2D pose for Probabilistic Robot Self Locator: We noticed the UKFPose2D may have some performance issues regarding Landmark Sensor Update, since it was generating unnecessary Landmark readings that would pollute the matrix calculations for pose provider, thus matrix calculations are more expensive we found a way to optimize it.
- Improve Image CNN Tools: As far as we saw, some Image CNN Tools like Image Transforms were overkill, difficult to understand and were making unnecessary calculations. We improved some of these issues by partially refactoring some of these tools.
Run this code
As we said, this code is heavily based in B-Human 2023 code release available in, therefore and so far, you can just follow their documentation to build and run the code.
Owner
- Name: Andres Santiago Santafe Silva
- Login: SantiagoSantafe
- Kind: user
- Repositories: 16
- Profile: https://github.com/SantiagoSantafe
GitHub Events
Total
- Watch event: 1
- Member event: 1
Last Year
- Watch event: 1
- Member event: 1