Sea Industrial PhD Program

An Initiative for the Best and the Brightest Global Talents (Open to All Nationalities)

Nurturing talent is one of SAIL’s core missions. With support from the Singapore Economic Development Board and the National University of Singapore (NUS), we are pleased to offer the Sea Industrial Postgraduate Programme (IPP) to promising individuals from all backgrounds who wish to pursue a relevant postgraduate degree whilst concurrently contributing to cutting-edge research at SAIL.

The Sea IPP gives prospective candidates the best of both worlds - a challenging and rewarding research career at SAIL as well as the opportunity to pursue world-class postgraduate qualifications at NUS. Sea IPP trainees undertake a jointly-identified research project supervised by both NUS and SAIL that contributes to their degree requirements.

Eligibility Criteria

The Sea IPP is open to global applicants who meet the following criteria:

  • Possess a strong interest in a career in computer science;
  • Meet our partner university's Ph.D. admission and academic requirements;
  • Obtained Bachelor's or Master's degree in Computer Science or related disciplines such as Engineering, Mathematics, or Physics;
  • Keen to study and perform research in the areas of Artificial Intelligence, Computer Vision, Deep Learning, Game AI, AI4Science, Natural Language Processing, Reinforcement Learning (Preferred but not limited to);
  • Not a recipient of, or hold any other scholarship or other awards of a monetary nature (Excluding bursary and financial aid);
  • For Sea IPP, applicants from all nationalities are welcome to apply; For EDB IPP, only Singapore Citizens and Permanent Residents are eligible to apply;
  • For applicants who are already enrolled in the Ph.D. program in the partner university, the candidature must be less than 2 years from the admission date & before passing the Qualifying Examination (QE).

Benefits

Sea IPP trainees will be able to enjoy the following benefits:

  • Competitive monthly salary including many employment benefits
  • Central Provident Fund contributions (for Singapore citizens and permanent residents only)
  • Medical Insurance, visa fee, etc.
  • Sponsorship of tuition fee and other university fees for 4 years
  • Conference and travel allowances
  • Computing equipment and resources

Application Process

To kickstart the Sea IPP application process, please complete the following steps:

Step 1: Submit an online application to NUS Graduate Admissions System. You are required to prepare your supporting documents as required by the partner universities. (For NUS, please refer to the SoC website for the latest requirement.)

Step 2: Once submitted the online application, please email the SAIL HR team (ipp.sail@sea.com) by quoting your NUS Application Number and attach the Research Statement and latest Curriculum Vitae.

Step 3: Shortlisted candidates will be invited for attending SAIL job interviews, in conjunction with the partner university's admission interview process. Both processes take approximately four to six weeks to complete, including document authenticity verification.

Step 4: Applicants may expect their application outcome by SAIL and NUS within two months from the date of application.

Application Deadline

The application is closed for the Intake 2022/23.

Other Information

If you have any queries regarding our Sea IPP, please contact our HR team (ipp.sail@sea.com).

Latest Publications

Inception Transformer
Chenyang Si, Weihao Yu, Pan Zhou, Yichen Zhou, Xinchao Wang, Shuicheng Yan
We present a novel and general-purpose Inception Transformer, or iFormer for short, that effectively learns comprehensive features with both high- and low-frequency information in visual data for tackling Transformer is incompetent in capturing high frequencies that predominantly convey local information.
Robustness and Accuracy Could Be Reconcilable by (Proper) Definition
Tianyu Pang, Min Lin, Xiao Yang, Jun Zhu, Shuicheng Yan
The trade-off between robustness and accuracy has been widely studied in the adversarial literature. Although still controversial, the prevailing view is that this trade-off is inherent, either empirically or theoretically. Thus, we dig for the origin of this trade-off in adversarial training and find that it may stem from the improperly defined robust error, which imposes an inductive bias of local invariance -- an overcorrection towards smoothness. Given this, we advocate employing local equivariance to describe the ideal behavior of a robust model, leading to a self-consistent robust error named SCORE. By definition, SCORE facilitates the reconciliation between robustness and accuracy, while still handling the worst-case uncertainty via robust optimization. By simply substituting KL divergence with variants of distance metrics, SCORE can be efficiently minimized. Empirically, our models achieve top-rank performance on RobustBench under AutoAttack. Besides, SCORE provides instructive insights for explaining the overfitting phenomenon and semantic input gradients observed on robust models.

We are hiring!