美图
Computer Vision Algorithm Intern (Image / Video Human Body Aesthetic Enhancement)
Computer Vision Algorithm Intern (Image / Video Human Body Aesthetic Enhancement)
发布于 大约 8 小时前实习/见习
澳大利亚
中级经验
实习生
仅现场办公
博士
实习与临时职位
PyTorch
Diffusion Models
Smpl
Human Pose Estimation
Video Generation
Multiview Generation
AI 估算 · 15k–25k
PhD-level internship in AI research at a listed tech company in Sydney; typical stipend range for top-tier research interns.
职位详情
关于这个职位
As a Computer Vision Algorithm Intern at Meitu's Sydney research center, you will work on cutting-edge 3D vision and human aesthetic enhancement technologies. You'll collaborate with a top-tier team, handling real-world datasets and state-of-the-art algorithms, with opportunities to produce both practical outcomes and high-quality academic publications. Ideal for PhD candidates passionate about AI-driven image/video enhancement and 3D human modelling.
最低要求
Experience:
· PhD candidate in Computer Science, Artificial Intelligence, Intelligent Arts, or related fields
· Proficient in Python, with experience using common libraries (e.g., NumPy, Transformers, Diffusers, Open3D), and strong hands-on experience with deep learning frameworks such as PyTorch
· Strong, sustained interest in image/video aesthetics or 3D art; resilient mindset and openness to change
Skills & Abilities:
· Solid engineering implementation skills and rigorous experimental practices; able to independently complete module-level tasks
· Proficient with AI-assisted programming tools; maintains clean, well-structured code
· Strong ability to read, analyse, summarize, and reproduce academic papers
· Clear logical thinking, strong problem-decomposition skills, and curiosity about underlying principles
Motivation & Values:
· Long-term interest in AI + imaging / 3D / aesthetics
· Proactive, self-driven, and willing to explore new approaches and challenges
· Alignment with a product culture centred on user experience and technical value
工作职责
· Research and implement algorithms related to depth estimation, multi-view generation, and 2D / 3D tracking with spatiotemporal reconstruction
· Follow state-of-the-art 3D vision papers and open-source projects; reproduce experiments and adapt methods to practical applications
· Collaborate with data teams to refine the 3D aesthetic development pipeline, improve data collection and quality evaluation, and establish foundations for high-quality scaling
· Explore the integration of human structure priors (Skeleton / SMPL / Mesh) with multi-modal cues such as depth, normals, and optical flow in reconstruction and generative models
· Assist in building data processing, evaluation, and visualization tools (e.g., immersive video aesthetic editing) to support rapid iteration
· Enable high-quality projection of 3D features into 2D visual outputs, with the goal of producing A-level or above academic publications
优先资格
Bonus Points:
· Prior experience with SMPL, human pose estimation, 3D reconstruction, or tracking
· Familiarity with video understanding, diffusion models, autoregressive models, or flow-matching methods
· Research projects, competitions, or open-source contributions
· Strong mathematical or geometric intuition; excellent aesthetic sensibility and passion for high-quality design
AI 洞察
优缺点分析
优点
- Work on cutting-edge 3D vision and generative AI with access to large-scale real-world data.
- Strong mentorship from a research-oriented team with a track record of publications.
- Located in Sydney CBD with potential for networking in a vibrant tech ecosystem.
- Opportunity to combine research and engineering, ideal for those aiming for both industry and academia.
- High expectations: you need to be self-driven and produce both code and papers within the internship period.
- Competitive field: 3D vision and generative AI demand strong mathematical and algorithmic foundations.
- Remote work not mentioned, requiring relocation to Sydney
- cost of living can be high.
- This role is perfect for PhD candidates in computer vision who enjoy hands-on engineering and have a passion for human aesthetics, 3D modelling, and generative models.
缺点 / 挑战
暂无明显挑战项
角色解读
- Transition into a full-time research scientist or engineer at Meitu or similar AI companies.
- Opportunity to publish at top conferences (CVPR, ICCV, NeurIPS) and establish academic reputation.
- Deepen expertise in 3D human modelling and generative AI, leading to specialized roles in AR/VR, gaming, or film production.
- Implement and adapt state-of-the-art 3D vision algorithms for human aesthetic enhancement and spatiotemporal consistency.
- Work with real product-scale datasets on tasks like depth estimation, multi-view generation, and 2D/3D human tracking.
- Collaborate with data teams to refine pipelines and build visualization tools for rapid iteration.
- Produce high-quality academic publications in parallel with applied research outcomes.
- Strong proficiency in Python and deep learning frameworks, especially PyTorch.
- Solid understanding of 3D vision fundamentals: reconstruction, tracking, and human body modelling (e.g., SMPL).
- Ability to read and reproduce academic papers
- rigorous experimental practices.
- Familiarity with generative models (diffusion, autoregressive) and multi-modal cues (depth, normals, optical flow).
申请策略
- Tailor your cover letter to express genuine interest in AI + aesthetics and alignment with Meitu's mission.
- Prepare a portfolio of relevant projects (GitHub) and be ready to discuss your problem-solving approach.
- Emphasize any prior research or projects in 3D reconstruction, human pose estimation, or generative models.
- Showcase publications, open-source contributions, or competition results in relevant areas.
- Highlight strong Python and PyTorch skills with examples of clean, well-structured code.
- Demonstrate aesthetic sensibility or art-related projects if available.
- Familiarize yourself with SMPL model and common 3D vision libraries (Open3D, PyTorch3D).
- Practice reproducing recent papers from CVPR/ICCV in areas like neural rendering or human digitization.
面试指南
- Use the STAR method: Situation, Task, Action, Result for project-based questions.
- For technical questions: explain the problem, outline possible approaches, justify your choice, and discuss trade-offs.
- Show enthusiasm for the domain: mention recent papers or tools you find exciting.
- Explain how you would approach fine-grained image aesthetic enhancement with controllable parameters.
- Describe your experience with 3D reconstruction pipelines
- what challenges did you face?
- How do you stay updated with the latest research in 3D vision? Give an example of a paper you implemented.
- What is your understanding of spatiotemporal consistency in video editing? How would you evaluate it?
职位点评
71
综合评分
Top-tier research internship at a global AI company, offering frontier 3D vision work and publication opportunities with modest pay.
更适合这类人
This role is ideal for candidates primarily motivated by skill development and cutting-edge research, who are less focused on immediate financial compensation.
表现最好
成长发展
相对薄弱
薪资福利
薪资福利50
成长发展95
工作生活60
使命价值80
薪资福利
50较低
As an internship, compensation is typically modest compared to full-time roles, and no specific benefits are mentioned. However, being in Sydney and at a listed company may offer reasonable stipend.
薪资信号未披露（AI估算：15K-25K/月）
成长发展
95较高
The role is heavily research-oriented, working on frontier technologies like 3D vision and generative AI, with strong potential for publications and skill growth.
技术前沿前沿/新兴技术
技术栈3D Vision、Diffusion Models、SMPL、PyTorch、Generative AI
成长机会publications、academic research、state-of-the-art
业务类型ambiguous
工作生活
60中等
On-site work in Sydney Central; typical office hours likely, but no explicit WLB signals. Relocation may be needed.
工作模式仅现场办公
办公地点市区核心地段
加班情况未提及（无法判断）
使命价值
80较高
The mission to 'Unite Art and Technology' and focus on aesthetic enhancement gives a sense of purpose, contributing to beauty and creativity in imaging.
行业发展高速增长赛道
社会影响正向社会影响力较高
使命信号Unite Art and Technology
创新程度积极采用新技术
Watch Jobs