Applied Digital Vision to Ovine and Porcine Production

Ref ABG-139955 Doctoral Thesis Subject 29/07/2026 Public/Private Funding

Laval University

Work Location

Quebec - Canada

Subject Title

Applied Digital Vision to Ovine and Porcine Production

Scientific Fields

Computer Science

Agronomy, Agri-food

Keywords

Digital Vision, AI, Deep Learning, 3D

Subject Description

Context:

Digital vision and artificial intelligence (AI) are already revolutionizing many sectors: healthcare, transportation, manufacturing industry. The agri-food sector represents today an exceptional field of application, where technological innovation responds to major societal challenges: food security, animal welfare, and sustainability of production.

The dairy sheep and pig production in Quebec are both facing urgent challenges: rapid detection of diseases, animal welfare monitoring, and labor shortages. Currently, most animal monitoring and inspections are performed visually by humans, which is time-consuming and prone to errors.

Digital vision, thanks to recent advances in deep learning, offers powerful solutions to automate these tasks and generate reliable information in real-time. This doctoral project is part of two complementary initiatives of the PaquetLab: the Ovision project, dedicated to the health and welfare of dairy sheep, and the 5-year Chair in Digital Vision Applied to the Pig Sector, carried out in partnership with major industry actors in Quebec (Quebec Pig Farmers, Olymel, CDPQ) and MAPAQ.

Doctoral Objectives:

These two 4-year doctoral projects aim to develop and validate robust digital vision algorithms:

In the context of the Ovision project in dairy sheep production: detect health problems early and assess the condition of the udder and the welfare of sheep from video and image flows.

In the context of the Chair in Digital Vision Applied to Pig Production and Processing: detect health problems in pigs early (respiratory diseases, abnormal behaviors) and automate the evaluation of animal condition, including the detection of lesions and anomalies on carcasses and viscera.

The student will work on real data (videos and images captured on farms, in clinics, and in commercial slaughterhouses), an opportunity rarely accessible in academic research, ensuring strong scientific and applied impact.

Scientific Challenges:

This project represents a unique opportunity to address cutting-edge issues in AI and digital vision, including: Object detection and multi-individual tracking in complex environments.

Behavioral analysis and animal posture recognition.

Image processing in uncontrolled conditions (variable lighting, occlusions, movement).

Detection of lesions and anomalies in biomedical/industrial imaging.

Development of robust and transferable models to real environments.

Use and development of 3D reconstruction techniques.

Start Date: 04/01/2027

Nature of Funding

Public/Private Funding

Funding Details

Public funding, but students are strongly encouraged to obtain scholarships for their training

Presentation of the Host Institution and Lab

Laval University

The PaquetLab is a dynamic and multidisciplinary research group at Laval University that brings together researchers from around the world to advance digital vision and artificial intelligence in animal sciences. Our mission is to leverage video, deep learning, and digital technologies to transform how we monitor, understand, and support animal production - whether in dairy sheep or pig production.

Led by Professor Éric Paquet, whose expertise accumulated over 25 years covers machine learning and data science in various domains, the laboratory develops tools that allow, in real-time, to detect health problems and track animal behavior directly on the farm. By transforming massive visual and sensor data flows into exploitable information, we aim to improve both animal welfare and production efficiency, in a sustainable and responsible approach.

We work closely with our industry partners to ensure that our innovations are not only scientifically rigorous but also practical, scalable, and impactful for producers. With a team that is both international and multidisciplinary, the PaquetLab is at the crossroads of agriculture, AI, and digital vision, contributing to building the new generation of technologies for intelligent agriculture.

Website: http://paquetlab.fsaa.ulaval.ca/

Doctoral Title

Doctorate in Computer Science

Country of Doctoral Degree

Canada

Institution Awarding the Doctorate

Laval University

Candidate Profile

Master's degree in computer science, computer engineering/software, applied mathematics, artificial intelligence, or related field.

Strong skills in deep learning (e.g., PyTorch/TensorFlow).

Experience in digital vision (3D, detection, classification, segmentation, object tracking).

Autonomy, creativity, and interest in concrete applications of AI.

Interest in agriculture or agri-food.

Application Deadline 30/09/2026