Project Description

Behaviour Model Lab specializes in behavior modeling and analysis using advanced technologies.

AQUAS: AI-driven Study of Fish Behavior

In the world of marine biology, understanding the intricate behaviors of fish has always been a difficult yet essential task. The ocean, with its complex ecosystems and myriad species, offers a challenging environment for scientists who seek to unravel the movements, interactions, and social behaviors of its inhabitants. Traditionally, researchers relied on manual observations, which were often time-consuming and limited in scope. However, thanks to advancements in artificial intelligence (AI), it is now possible to track and analyze the behavior of fish more efficiently and with greater accuracy.

At the AQUAS, scientists have embraced the power of AI to study the behavior of 14 distinct fish species in a controlled aquatic environment. By employing machine learning algorithms and real-time monitoring, the AI system not only tracks the movements of these fish but also learns to differentiate between species, providing unique insights into their behavior, interactions, and ecological roles. With only 50,000 lines of code, the AI is able to observe, record, and analyze vast amounts of data, ultimately helping researchers better understand these fascinating creatures.

The Fish Species Under Study

The AQUAS focuses on 14 species of ocean fish, carefully selected for their varying behaviors, appearances, and roles within their ecosystems. Each species is studied for its movements, social interactions, and responses to environmental stimuli. Here is a brief overview of each species:

Clownfish (Amphiprioninae)

Description: Clownfish are small, brightly colored fish with distinct orange bodies and white stripes. They are famous for their symbiotic relationship with sea anemones, which provide them with protection while the clownfish defend the anemones from predators.
Behavior Focus: The AI tracks their interactions with sea anemones and other species, monitoring their territorial behavior and social structure within anemone groups.

Lionfish (Pterois)

Description: The lionfish is an exotic, venomous fish known for its striking appearance, featuring long, spiny fins and a fan-like tail. Native to the Indo-Pacific, it has become an invasive species in the Atlantic.
Behavior Focus: The AI observes their predatory behavior, including how they hunt smaller fish and their interactions with other fish species in the environment.

Blue Tang (Paracanthurus hepatus)

Description: Also known as the "Regal Tang," this bright blue fish with a yellow tail is famous for its role in Finding Nemo. It is often found in coral reefs, where it plays a key role in maintaining reef health.
Behavior Focus: The AI studies its schooling behavior, interactions with other reef species, and its movement patterns within the coral ecosystem.

Damselfish (Pomacentridae)

Description: Small and often brightly colored, damselfish are known for their aggressive territorial nature. They are found in tropical and subtropical marine environments, particularly in coral reefs.
Behavior Focus: The AI tracks their territorial defense, school dynamics, and interactions with other species, especially in relation to their defense of feeding territories.

Cardinalfish (Apogonidae)

Description: Cardinalfish are small, nocturnal species often found in coral reefs and seagrass beds. They are characterized by their large eyes and often reddish or silvery coloring.
Behavior Focus: The AI focuses on their nocturnal behaviors, how they school for protection, and their responses to environmental changes such as light or predator presence.

Pygmy Angelfish (Centropyge)

Description: These small, brightly colored fish are often found in reef environments. They are known for their vibrant hues, including shades of blue, yellow, and orange.
Behavior Focus: The AI tracks their movements within the reef, focusing on their feeding patterns and social interactions within small groups or with other species.

Neon Tetra (Paracheirodon innesi)

Description: A small freshwater fish known for its glowing blue and red coloration, the neon tetra is often found in schools. Although native to South American rivers, it can be found in brackish coastal waters.
Behavior Focus: The AI observes their schooling behavior and how they interact with other species in the ecosystem, as well as their responses to environmental stimuli such as changes in water conditions.

AI System and Data Collection

The AI system within the AQUAS is designed to track the movements of these 14 species in real time. Using high-definition cameras, motion sensors, and advanced machine learning algorithms, the system is capable of identifying each fish by its unique physical characteristics and behavioral patterns. The AI then processes this data to generate detailed movement trajectories, social interactions, and responses to environmental changes.

Data Collection and Tracking:

Each fish is equipped with a tracking system that enables the AI to monitor its position within the tank or natural habitat. The system can distinguish between species, even those with similar physical appearances, such as the clownfish and damselfish.

Movement Analysis:

The AI uses algorithms to analyze swimming patterns, body movements, and social behavior. It tracks how fish move within their environments, whether they are swimming alone or in schools, and how they interact with other species.

Environmental Response:

By introducing controlled changes in the environment, such as changes in water temperature or light intensity, the AI can monitor how the fish species respond. It learns to predict how certain species will react to environmental shifts, providing insights into their adaptability.

Learning and Differentiation:

Over time, the AI learns to distinguish not only between species but also between individual fish. This learning process enables the system to build more detailed models of behavior, allowing researchers to better understand each fish’s unique traits and how they contribute to the broader ecosystem.

Insights Gained from AI Observations

The use of AI in the AQUAS has already provided valuable insights into fish behavior, including:

• Species Interactions: By tracking how different species interact, the AI has helped identify patterns of cooperation, aggression, and territoriality. For example, damselfish and clownfish both defend their territories, but the AI has observed that clownfish are more likely to share their habitat with other species.
• Feeding Behavior: The AI has also shed light on feeding patterns, showing how certain species, like the blue tang and pygmy angelfish, forage for algae, while others, like the lionfish, actively hunt smaller fish.
• Social Structures: The AI system has revealed how different species organize themselves within schools. Species like the cardinalfish and neon tetra form tight-knit groups for protection, while others, like the lionfish, tend to be solitary.
• Environmental Adaptation: The AI has observed how these fish species adapt to environmental changes, such as shifts in water temperature or light levels. Some species, like the clownfish, are more resilient to changes in habitat, while others are more sensitive.

Conclusion

The AI-powered AQUAS represents a major leap forward in understanding fish behavior. By tracking the movements and interactions of 14 distinct species in a controlled environment, the lab is able to gather insights that would be nearly impossible to obtain through traditional research methods. With the help of advanced machine learning algorithms, scientists can now study fish behavior in greater depth, uncovering patterns and relationships that will ultimately help to protect and preserve these vital marine species. The data collected by the lab will not only further our understanding of aquatic ecosystems but also contribute to the broader field of marine biology, providing crucial information for conservation efforts and the management of ocean habitats.

Additional Information

Here you can place additional information, such as contacts, documents, or useful links.