Perception and Recognition Models in Competition
This simulation pits different perception and recognition models against each other in direct competition. Observe how various strategies perform under different conditions and which combinations of genes prove most advantageous for survival.
Competition
It is important to notice that the cells of each population interact with each other and one winning over the other isn't simply one proliferating out the other. For an explanation of how perception and recognition is implemented see:
Comparing a perception model to a recognition model is unfair, as the fitness gained from the total perception interactions is significantly higher. However, comparing two models within the perception group or the recognition group is fair.
Cell Attributes and Life Cycle
Cell Properties
- Four Circular Genes (A, B, C, D, Z)
- Fitness (0-255)
- Age (+1 with each cycle)
Life Cycle
- Fitness +1 for each neighbouring empty cell.
- Fitness gain or loss from each neighbour depending on selected model and corresponding gene comparison matches.
- At fitness >= 255 fitness is reduced to 255 minus random value in between 0 and 60.
- At fitness >= 128 cell tries to reproduce, dividing fitness equally to offspring and parent.
- Reproduction may fail if several cells try to divide into an empty cell. Winner is the offspring with the highest fitness.
- During reproduction, there is a 1 in 10,000 chance that a cell mutates a gene, with a mutation range of ±64.
- At fitness <= 0 or age > 40, the cell dies.
All content on this page — including photos, videos, and field audio recordings — was created by Karl-Heinz Müller, a Montréal-based wildlife filmmaker, sound recordist, and photographer. Founder of MUUUH and Québec Sauvage, he has spent over a decade documenting Canada’s natural habitats through immersive soundscapes and wildlife films. Learn more about Karl-Heinz
Last updated: 2025-11-14