Research Intern – Human-Centric 3D Generative Models - Jobs bei Apple (CH)

Research Intern – Human-Centric 3D Generative Models

Zurich, Zurich, Switzerland

Students

At Apple, we push the boundary of what is possible to improve the lives of our customers. The Zurich Vision Lab (ZVL) is focused on improving communication by advancing the state-of-the-art in computer vision and machine learning. We are shipping features such as Persona for Apple Vision Pro, Portrait Mode, ARKit, Animoji and FaceTime Eye Contact. These breakthroughs are enabled by state-of-the-art research combined with world-class design. We are looking for a research intern who is excited to contribute to the next generation of Apple products. You will be working on large-scale 3D generative models for human representation.

Description

This position requires a highly motivated individual who wants to join us in the fast-moving field of large-scale generative models for generation of human-centric 3D representations.You will be responsible for developing, implementing, evaluating, and improving algorithms that model humans in complex and diverse scenarios in a photorealistic and 3D-consistent way.

Minimum Qualifications

• Strong understanding of generative models and/or multi-modal foundation models.
• Strong understanding of neural rendering, 3D representations and Gaussian Splatting.
• Fluency in Python and Machine Learning Frameworks (i.e. PyTorch) is required.
• Fluency in English is required.
• Excellent communication and collaboration skills.
• Demonstrable problem solving and analytical thinking abilities.
• Currently enrolled in PhD or Masters degree in Computer Science or equivalent. PhD degree paths are preferred for research focused internships.
• (Please note: At the end of the internship, you must return to university to continue your education, or the internship must be the last requirement for you to graduate.)

Key Qualifications

Preferred Qualifications

• Availability for 6 months minimum is preferred.
• Hands-on experience with 3D generative models, gaussian splatting, or vision foundation models.
• Publication record at top conferences such as CVPR, ICCV, ECCV, ICLR, NeurIPS or SIGGRAPH is a plus.