Paquet lab

AI and computer vision for the livestock sector

Chronobiology, genomics and systems biology

Research interests:

Our research is at the intersection of animal genomics, bioinformatics, artificial intelligence, computer vision, sensor development, and precision livestock farming. We are passionate about transforming complex biological and production data into actionable knowledge that improves animal health, welfare, productivity, and sustainability. Through our MAPAQ Research Chair in Digital Vision for Smart Animal Production, we develop innovative imaging technologies, intelligent sensors, and machine learning approaches to monitor animals and their environment in real time. Our work integrates data from genomic analyses, automated phenotyping systems, wearable and environmental sensors, and computer vision platforms to better understand animal physiology and behavior. A major focus of our research is the development of next-generation decision-support tools capable of combining large-scale genomic information with continuous streams of data generated on modern farms. By leveraging advances in artificial intelligence and data science, we aim to accelerate genetic improvement, optimize management practices, and support sustainable livestock production systems. We are also exploring emerging questions in systems biology and chronobiology to better understand how biological rhythms influence animal health, metabolism, reproduction, welfare, and performance. These novel perspectives open exciting opportunities for developing innovative management strategies adapted to the biological needs of animals. Our laboratory offers students a highly interdisciplinary environment where biology, engineering, computer science, and data analytics converge. We welcome students interested in developing cutting-edge technologies and data-driven solutions that will shape the future of animal production.

In collaboration with leading national and international research partners, our team is also developing innovative machine learning and deep learning models based on milk mid-infrared (MIR) spectroscopy data. These projects aim to unlock the full potential of routinely collected milk spectra for the early detection of health and metabolic disorders, the monitoring of animal resilience and welfare, and the prediction of processing and technological properties of milk. By combining MIR phenotypes with genomic, sensor, environmental, and production data, we seek to develop robust predictive tools that will support precision management on farms and enhance the value chain from animal production to dairy processing.