Postdoctoral Research Visit in Generalisable Early Classification of Bovine Embryo Morphokinetic Profiles from Time-Lapse Videomicroscopy in Inria Centre at Rennes University (MALT Team) – France

Summary

The Inria Centre at Rennes University (MALT Team) offers a two‑year fixed‑term postdoctoral position to develop real‑time deep‑learning tools for early classification of bovine embryo morphokinetic profiles from time‑lapse video. The project aims to create online classifiers that predict embryo transferability before day four, assess earliness‑reliability trade‑offs, improve robustness to domain shift, and evaluate generalisation on other species. The role includes delivering reproducible software and publishing results. Candidates must hold a PhD in computer science or a related field, have strong deep‑learning expertise (especially with PyTorch), and possess a solid experimental methodology. The position provides a monthly gross salary from €2,788, generous leave, subsidised meals, transport reimbursement, teleworking after six months, and professional equipment. No funding is secured yet, but BIENVENÜE funding may be pursued.

Key Facts

Department	MALT Team
Position Type	Postdoctoral Researcher
Discipline	Computer Science
Research Area	Machine Learning, Computer Vision, Biomedical Imaging
Degree Required	PhD degree
Salary	Monthly gross salary from €2,788
Funding	No funding is secured yet
Contract Duration	Fixed-term, 2 years
Language	English
Deadline	2027-03-31
Required Documents	resume, cover letter, letters of recommendation

About the Project

The project will develop deep‑learning based, real‑time analysis tools that classify bovine embryos from raw time‑lapse video into a detailed taxonomy of transferability profiles as early as possible, ideally before four days of development. It will design early‑decision classifiers, quantify earliness‑reliability trade‑offs, ensure robustness to domain shifts, and test generalisation on embryo datasets from other species, delivering reproducible software and scientific publications.

Key Responsibilities

• Design early and online classification methods processing video streams incrementally with confidence estimates
• Quantify earliness‑reliability trade‑off and define evaluation protocols to identify earliest reliable recognition time
• Assess and improve robustness to domain shift using data augmentation, domain adaptation, or self‑supervised pre‑training
• Evaluate methods on publicly available embryo datasets from other species to assess generalisation
• Deliver reproducible, documented software together with scientific publications

Required Skills

• PhD in computer science, applied mathematics, machine learning/computer vision, or related field
• Strong expertise in deep learning for video or image sequences (CNNs, transformers, temporal models) with solid PyTorch practice
• Sound experimental methodology: evaluation design, handling class imbalance, small/heterogeneous datasets, statistical rigour
• Proven scientific track record with publications in relevant venues
• Ability to write clearly and communicate in English (French a bonus)
• Experience in early time‑series classification, domain adaptation, self‑supervised learning, biomedical or microscopy data, or developmental biology is appreciated
• Good software engineering practice (version control, reproducibility, code documentation)
• Autonomy and ability to work in an interdisciplinary team and supervise students

Who Should Apply

Applicants with a PhD in computer science, applied mathematics, machine learning, computer vision or a related discipline, who have strong deep‑learning experience for video or image sequences, solid experimental methodology, a proven publication record, and the ability to communicate in English (French a plus) are encouraged to apply.

Benefits

• Monthly gross salary from €2,788
• 7 weeks of annual leave plus 10 extra RTT days
• Subsidized meals and partial reimbursement of public transport costs
• Teleworking possible after 6 months with flexible working hours
• Professional equipment available (videoconferencing, loan of computer equipment)
• Access to social, cultural, and sports events and activities
• Access to vocational training

About the Employer

The Inria Centre at Rennes University is one of Inria's eight centres, hosting more than thirty research teams and situated within a rich R&D and innovation ecosystem.

Frequently Asked Questions

What is the application deadline?

The deadline is March 31, 2027.

What documents must be submitted?

Applicants must submit a resume, cover letter, and, if applicable, letters of recommendation via the Inria website.

Is the position funded?

No funding is secured yet, but applicants may apply for BIENVENÜE funding during fall 2026.