Machine Learning Engineer, R&D
CLO Virtual Fashionover 3 years ago
Seoul, Korea, SouthMid Level
Responsibilities
- Develop and improve 3D polygon mesh estimation models from varied inputs including images, prompts, and video.
- Develop and improve parametric CAD model estimation models from varied inputs.
- Develop and improve 3D avatar pose and shape estimation models from varied inputs.
- Fine-tune, optimize, and serve large generative models such as Stable Diffusion for specific purposes.
- Improve the efficiency of other complex tasks by applying deep-learning models.
Requirements
- Hold a master's or doctoral degree in deep-learning-based vision or graphics research, or have at least two years of equivalent industry experience.
- Have research project experience in vision, 3D deep learning, or generative models.
- Be proficient in Python and PyTorch.
- Be motivated to learn and grow collaboratively.
- Preferred qualifications include experience with mesh or graph deep-learning models, diffusion or StyleGAN models, computer graphics, geometry processing, physically based rendering, C++ service or application development, Dockerized API services, or ONNX-based on-device services.
Benefits
- Dual monitors and a desktop, laptop, MacBook, or iMac, with equipment upgrades every three years.
- Professional-development and fitness costs, book purchases, current tools, and club support.
- Breakfast salad bar, lunch, snacks, and dinner meal support when working late.
- Legal annual leave, bereavement leave, health-examination leave, and year-end holiday from December 25 through January 1.
- Flexible working hours with arrival between 7:00 and 10:00 AM and departure between 4:00 and 7:00 PM.
- Health examinations at major hospitals, health-examination leave, bereavement support, and statutory social insurance.
- Flexible vacation use without a separate approval process when coordinated with teammates.
- Support for one month of work from the Jeju office, including accommodation and round-trip airfare for family stays.
- Slack-based global communication, quarterly knowledge sharing, and a non-hierarchical work culture.