I'm a research scientist at INRAE. Working in the MIA & the BIOEPAR research labs, where I am currently doing my PhD (the PACE-COW project).

My research explores how artificial intelligence can help us better understand and care for animals. I focus on developing tools that automatically detect and quantify cattle behaviour using video, accelerometers, and proximity sensors. By combining computer vision, deep learning, and ethology, we try to detect diseases before they become visible to the human eye.

Within the PACE-COW project, I design multi-sensor pipelines that synchronize large volumes of barn data to reconstruct the daily life of dairy cows. This includes :
- designing and testing methods for annotating large amount of video data (detection, tracking and classification of cows in videos) and validating them against real observations
- training models for behaviour classification from accelerometer data (Time Series Classification)
- analysing how time budgets relate to animal welfare and health.

My broader goal is to bridge the gap between AI research and practical farming, making Precision Livestock Farming technologies both scientifically rigorous and genuinely useful for farmers, veterinarians, and welfare scientists.

Research

I'm interested in computer vision, time-series, deep learning, and multimodal AI for animal behaviour analysis. Most of my research focuses on inferring cattle activity and welfare from sensor data.

Publications

You can also find my publications on Google Scholar and HAL.