Research Scientist Graduate (3D/4D Reconstruction/...

As a technology brand with independent innovation and R&D capabilities, PICO is committed to becoming a leading XR platform, helping developers and creators achieve success, and jointly creating a better life experience for global consumers. It has pioneered the expansion of virtual reality into multiple fields, bringing new experiences to consumer-grade scenarios such as sports, video, and entertainment, and is widely used in commercial scenarios such as education, medical care, and corporate training, providing diversified VR solutions for global enterprise users. We are looking for talented individuals to join our team in 2026. As a graduate, you will get opportunities to pursue bold ideas, tackle complex challenges, and unlock limitless growth. Launch your career where inspiration is infinite at ByteDance Successful candidates must be able to commit to an onboarding date by end of year 2026. Please state your availability and graduation date clearly in your resume. Responsibilities 1. Participate in research and development of 3D and 4D scene reconstruction/generation algorithms (including dynamic scenes and human subjects), and/or lighting estimation/editing, and promote their application in MR/XR scenarios; 2. Develop efficient and scalable 3D/4D reconstruction/generation/relighting pipelines, including task formulation, dataset construction, model training and optimization, inference acceleration, and deployment; 3. Collaborate closely with cross-functional teams at PICO to promote hardware-software co-innovation; 4. Track the latest research progress in 3D vision and computer graphics, and contribute to papers and patents. Minimum Qualifications - Individuals who are completing or have recently completed a PhD degree in Software Development, Computer Science, Computer Engineering, or a related technical discipline. - Excellent programming skills, proficient in C++ or Python; - Solid foundation in 3D Computer Vision and/or Computer Graphics, including: - Multi-view reconstruction (e.g., MVS, LRM);3D/4D representations (e.g., NeRF, Gaussian Splatting); -Familiarity with ray tracing and rasterization-based offline or real-time rendering algorithms is a plus; - Knowledge in rendering is a plus, including fundamentals, digital asset creation and analysis (e.g. Blender, UE, Mitsuba), Image-based rendering and lighting (IBR, IBL), and lighting estimation. Preferred Qualifications - Experience or deep understanding of NeRF or 3D Gaussian Splatting (3DGS), including: - Familiarity with 3DGS variants such as Mip-Splatting and 3DGS-MCMC; - Understanding of dynamic Gaussian, 4DGS reconstruction, volumetric video, and feed-forward 3D/4D generation algorithms; - Research or project experience in high-fidelity, efficient training, or compact 3D/4D representation. - Familiarity with downstream applications of 3D/4DGS, including:- GS Relighting (e.g., GaussianShader, GIR, IntrinsicAvatar);GS Editing (e.g., Gaussian Grouping, GaussianEditor). Knowledge of inverse rendering and lighting estimation, including learning- and optimization-based models such as Differentiable Path Tracing and Physically-based Inverse Rendering; - Familiarity with computational imaging pipelines, including sensor fusion, exposure bracketing, tone mapping, and color management for mobile camera systems - Publications in accredited conferences or journals (CVPR, ICCV, SIGGRAPH, NeurIPS, etc.) are a plus, - - Winners or high performers in programming or algorithmic competitions such as ACM-ICPC, NOI/IOI, or TopCoder are highly preferred.

Created: 2026-04-15