Job Summary: We are seeking a professional for an R&D project focused on structured image generation, emphasizing controllability, visual consistency, and identity persistence. Key Highlights: 1. Research and Development (R&D) project in image generation. 2. Focus on controllability, visual consistency, and identity persistence. 3. Opportunity to work with advanced generative models. ***Role Summary*** This position is for a Research and Development (R&D) project focused on **structured image generation**. The goal is to identify a candidate with expertise in **controllability, visual consistency, and identity persistence**. The ideal candidate must have hands-on experience with generative models, model adaptation, multimodal conditioning, and technical *benchmarking*. ***Main Responsibilities*** * Investigate advanced methods for structured image generation. * Develop strategies focused on structural conditioning. * Work on creating and improving mechanisms for visual identity persistence (e.g., character/object consistency). * Explore intermediate representations to achieve greater control over image generation. * Conduct and analyze comparative *benchmarks* across different technical approaches. * Support technical validation and rigorous documentation of experiments. ***Mandatory Requirements*** * Bachelor’s or higher degree in Computer Engineering, Artificial Intelligence, *Machine Learning*, or a related field. * Solid, proven experience with generative image models. * Practical knowledge of model adaptation techniques (*fine-tuning*, *adapters*, *embeddings*, LoRA, etc.). * Proficiency in Python programming and major *Machine Learning* frameworks. * Demonstrated ability to design experiments, objectively compare results, and interpret metrics. * Strong technical foundation combined with rigorous analytical thinking. ***Preferred Qualifications (Highly Valued)*** * Prior experience with *diffusion models* and *vision-language models*. * Previous specific work in character consistency, visual identity, or structural control. * Active and relevant presence on GitHub, academic/scientific publications, or a strong technical portfolio in this domain. ***How to Apply*** Interested candidates must submit their **CV**, along with links to their **GitHub**, relevant **publications**, or **technical projects** demonstrating experience in this field. **Location:** Cascais (On-site) **Position Type:** Full-time | R&D Project (24-month contract) **Annual Salary Range:** €35,000 to €65,000 (adjusted according to the candidate’s actual seniority) Employment Type: Full-time Compensation: 35 000,00€ – 65 000,00€ per year
João Santos
Indeed · HR
João Santos
Indeed · HR
