In the ever-evolving intersection of neuroscience and artificial intelligence, a team of researchers from Nagoya University, in collaboration with Osaka University and Tohoku University, has unveiled a groundbreaking AI system capable of identifying and manipulating animal behaviors with unprecedented precision. This cutting-edge technology, named YORU (Your Optimal Recognition Utility), is poised to revolutionize how scientists study social interactions in animals by providing real-time behavior recognition coupled with targeted neural control.
The compelling story begins within the confines of a laboratory chamber housing fruit flies. A male fly, engaged in a quintessential courtship ritual, extends and vibrates his wings, producing the species-specific love song meant to entice a nearby female. Unexpectedly, a fleeting green flash interrupts the serenade, causing the male to halt mid-note and fold his wings. This interruption isn’t the result of a random event but a precise intervention by YORU, which detected the onset of the courtship behavior and instantaneously activated a neural silencing mechanism. The female fly, witnessing the interrupted performance, promptly disperses, illustrating the real-time behavioral impact achieved through this technology.
YORU’s fundamental innovation lies in its ability to monitor groups of animals and distinguish individual behaviors with remarkable accuracy. Traditional techniques, such as frame-by-frame tracking of specific body parts, are cumbersome and often ineffective when animals overlap or interact closely. YORU, contrastingly, employs an object-based approach to analyze entire behavioral patterns within a single video frame, bypassing the need for complex motion tracking and enabling rapid real-time processing. This method has demonstrated a high degree of accuracy, detecting behaviors across distinct species—ranging from ants sharing food and zebrafish showing social orientation to mice engaged in grooming—with success rates exceeding 90%.
One of the pivotal advantages of YORU is its exceptional speed and efficiency. The AI system processes visual data approximately 30% faster than contemporaneous tools, a critical improvement for experiments where split-second timing governs the validity of results. This speed, coupled with its object-based recognition strategy, circumvents challenges expounded by occlusion or multiple interacting subjects, which traditionally hinder behavior detection algorithms. Consequently, YORU enables continuous operation and closed-loop feedback during animal interactions, marking a new frontier in behavioral neuroscience.
The real magic unfolds when YORU interfaces with optogenetics—a revolutionary technique that employs light to control genetically modified neurons. Through a triadic process involving genetic engineering, behavioral detection, and light-mediated neuronal control, the research team achieved a system capable of shutting down specific neural circuits instantaneously as targeted behaviors arise. Firstly, animals are genetically engineered to express opsins, specialized light-sensitive proteins, in select brain neurons. These proteins act as molecular switches that respond to precise wavelengths of light, enabling activation or inhibition of neurons on demand.
Once animals are equipped with opsins, YORU continuously monitors their behavior using high-resolution cameras. When the AI detects a pre-defined behavior—in the fruit fly experiment, the extension of wings initiating courtship—it immediately sends an electrical signal to activate a localized light source. This light is meticulously directed at the target animal, ensuring other members of the group remain unaffected. This achievement surmounts a significant obstacle in neuroscience research, where previous methods illuminated entire experimental arenas, confounding individual-specific behavioral analysis.
Upon illumination, the opsins within the target neurons respond by opening ion channels in the neural membranes, resulting in either activation or silencing of those neurons. This precise control alters the brain activity linked to the detected behavior, enabling researchers to examine causal relationships between neural circuits and social actions in real time. The impact on behavior is immediate and measurable, as evidenced by the cessation of the male fruit fly’s song in response to neuron silencing.
YORU extends beyond merely observing behavior; it is a transformative platform for manipulating animal social dynamics. By selectively influencing single individuals within a social group, neuroscientists can dissect the nuanced roles and contributions each animal plays in complex behaviors. This capability holds profound implications for understanding the neural basis of social cognition, communication, and interaction across species.
Moreover, the versatility of YORU is amplified by its adaptability and user accessibility. Scientists can train the system to recognize novel behaviors with minimal datasets and without any requirement for programming expertise. This democratization of advanced behavior detection and control tools promotes widespread adoption, accelerating discoveries in sensory neuroscience, ethology, and behavioral genetics.
To illustrate its potential, the research team demonstrated YORU’s applicability on diverse species besides fruit flies, including ants displaying cooperative food sharing, zebrafish demonstrating social orientation, and grooming behaviors in mice. These results highlight the AI’s broad scope and the possibility of comparative neuroscience studies across taxa.
The broader implications of this research are multifaceted. By shedding light on real-time cause-and-effect relationships between neural activity and behavior, YORU empowers researchers to probe fundamental questions about the brain’s orchestration of social functions. It also lays the groundwork for developing novel interventions for behavioral disorders and enhancing AI-human-animal interactions.
Importantly, the research team has made YORU accessible to the global scientific community via an online platform, fostering collaborative exploration and ensuring transparency. This open-access approach promises to accelerate the pace of neuroscience research, inspire innovations, and facilitate interdisciplinary partnerships.
In summary, YORU represents a monumental leap in the study of animal behavior and neuroscience, merging sophisticated AI-driven detection with precise optogenetic control. This synergistic methodology deciphers the brain-behavior nexus within the dynamic context of social interactions, offering a powerful toolset to unravel the complexities of animal cognition with unmatched temporal and spatial resolution.
Subject of Research: Animals
Article Title: YORU: Animal behavior detection with object-based approach for real-time closed-loop feedback
News Publication Date: 11-Feb-2026
Web References:
https://www.science.org/doi/10.1126/sciadv.adw2109
References:
Yamanouchi et al., 2026. “YORU: animal behavior detection with object-based approach for real-time closed-loop feedback.” Science Advances, DOI: 10.1126/sciadv.adw2109
Image Credits:
Yamanouchi et al., 2026
Keywords:
Artificial Intelligence, Animal Behavior, Optogenetics, Neural Control, Real-Time Feedback, Social Interaction, Neuroscience, Computer Vision, Genetic Engineering, Fruit Fly, YORU, Behavioral Neuroscience
