Visual Generation & Multimodal Evaluation Researcher Graduate (AML-Ark-US) - 2027 Start (PhD)

What you'll do

• Build evaluation systems for image and video models and agents, covering quality, instruction following, multimodal understanding, and safety.
• Develop automated metrics, model-based evaluators, and reproducible human evaluation protocols; create video generation and debugging agents.
• Build large-scale image and video data pipelines and translate evaluation findings into model and product improvements.

What they ask for

• Completing or recently completed a PhD in Computer Science, Artificial Intelligence, Machine Learning, Computer Vision, or a related field.
• Solid foundation in deep learning and computer vision, including generative modeling fundamentals.
• Practical experience in at least one of visual generation, multimodal LLMs, video understanding, or visual quality assessment.
• Strong Python skills and proficiency with PyTorch or an equivalent framework, or a multimodal evaluation framework; demonstrated research or engineering ability.

Nice to have

• Publications at top-tier vision or ML venues, such as NeurIPS, ICML, CVPR, ICCV, or ECCV.
• Experience with diffusion-based models and post-training, visual generation benchmarks or evaluation frameworks, agent frameworks for creative workflows, or large-scale video data infrastructure.

Tools

Python, PyTorch

AI focus

Image and video generation evaluation, multimodal understanding, model-based evaluation, and agents that orchestrate video-generation workflows.