JOB DESCRIPTION
DataLab is inviting applications for a full-time postdoctoral scholar in image processing and classification beginning Fall 2020. This postdoc will contribute independent research to a Getty funded project focused on developing software for classifying digital images of historical printed materials based on both semantic and visual similarity. The scholar will develop improved segmentation, feature extraction, and statistical image analysis and modeling methods, and develop and implement both trained (CNN) and untrained (Deep Learning) classification systems. The scholar will work closely with a multi-disciplinary team of staff and faculty members at DataLab in the performance of this research and to integrate developed methods into a functional, open source software package that is being developed by the team.
We seek applicants with a keen interest, proven research record, and experience grappling with problems of image classification. Specifically, the applicant should have interest in developing machine learning methods for correctly classifying images that are visually dissimilar but semantically similar. For example, an image of a crown and an image of a throne are visually dissimilar but are both about “royalty.” Similarly, a drawing of Buckingham Palace and a painting of Buckingham Palace are visually dissimilar, but both depict the same thing. The scholar will have access to large collections of both labeled and unlabeled images as well as contextual text descriptions for many images. Beyond approaching the theoretical problems that underly this classification problem, the candidate must also be committed to developing applied methods for solving these problems and integrating these into a collaboratively developed software platform.
The candidate will work as part of a highly collaborative and interdisciplinary team of image, natural language processing experts as well as humanities and data science scholars. The position requires both scholarly independence and the ability to manage scientific projects and the ability to integrate into a larger team and community. This is an excellent opportunity for individuals who want to gain experience in a broad array of authentic data science methods, and who are keen to learn new technologies/methods and apply them in novel ways to solve real problems.
DataLab’s mission is to promote and support research and training in data-driven discovery across all colleges, schools, and disciplines at UC Davis. The DataLab coordinates data-science training activities, provides consulting and collaborative services for research projects, and conducts novel research in data science. Research and training activities at DataLab span the entire data pipeline, from data acquisition, management, cleaning, transformation, visualization, and modeling to dissemination/publication and involve aspects of high-performance computing, data governance, security, privacy and ethics. DataLab is committed to fostering an inclusive environment to promote all members of the university community, including faculty, students and staff from a variety of domains, backgrounds, cultures and personal experiences. For more information, see https://www.ucdavis.edu/ and http://datalab.ucdavis.edu.
QUALIFICATIONS
Applicants for the position should have earned a PhD prior to the position start date and have both interest and experience in Data Science with the following background in image processing and classification:
- a PhD in a field related to Data Science (i.e., Computer Science, Statistics, Machine Learning, Information Science, etc.) with a proven ability to apply methods and technologies to solve cutting-edge problems in different domains, and
- a thesis that reflects doctoral research in the area of image processing and classification as well as machine learning;
- demonstrable proficiency in the areas of image processing and classification and machine learning.
All applicants are expected to have proficiency in a high-level programming language (e.g., C, R or Python) and will be expected to contribute C++, R, and PHP code to the project codebase. The ideal candidate will have demonstrated strong project leadership and management, interpersonal communication, and decision-making skills as well as the ability to work well independently and as part of a team. Motivation and demonstrated capacity to pursue complex problems and work successfully within a collaborative setting are also essential. Preferred candidates will also have some/any of the following skills: use of SQL databases; high-performance computing; and web/JavaScript-based data visualization tools.
MAJOR RESPONSIBILITIES
The postdoctoral scholar is expected to:
- Work both independently and in collaboration with DataLab staff and research affiliates to develop novel image classification methods;
- Work with DataLab staff to integrate novel methods into open source application codebase;
- Document developed methods and software;
- Participate in quality assurance and testing activities;
- Publish research findings independently and in collaboration with other DataLab faculty, staff, and affiliates;
- Participate in regular lab and project meetings;
- Work on-site in DataLab’s collaboratory.
Training and opportunities for the postdoctoral scholar include:
- Dedicated research mentoring;
- Research funding to support approved travel and dissemination of research conducted at DataLab;
- Access to the computing resources at UC Davis and DataLab, including compute servers and clusters as well as Amazon Web Services resources;
- Access to opportunities for specialized training sponsored by DataLab and other scientific computing groups at the University;
- Teaching opportunities in workshops for relevant areas of specialty and possible teaching of undergraduate data science courses;
- Opportunity to mentor graduate and undergraduate students working on the research project;
- Contribute to other DataLab research endeavors as appropriate according to personal development plan and interests.
The postdoctoral fellow will report to the Executive Director of the DataLab. This is not a tenure-track position and is for one year, with the possibility of extension. Salary will be commensurate with experience and university policy. Given the sensitive nature of the data involved in research at the DataLab, this position is subject to background check. Employment is contingent upon successful completion of a background investigation including criminal history and identity check. The successful applicant must be eligible to work in the USA.
UC Davis is an affirmative action/equal employment opportunity employer and is dedicated to recruiting a diverse community. We welcome all qualified applicants to apply, including women, minorities, veterans and individuals with disabilities.
For additional information, contact datalab@ucdavis.edu.
HOW TO APPLY
Documents: To apply please submit the following items as one pdf document:
- Curriculum Vitae, including listing of pertinent coursework and relevant code repository (Github, Bitbucket, etc.);
- Cover Letter, including your desired start date and names of your 3 references;
- Statement of Research Background and Demonstrated Competency in Image Classification and Machine Learning
The following may optionally be included:
- Brief Example Research Plan, including research interests that align with project (see description at here) and learning strategy for skills that need development.
- Transcripts from graduate and undergraduate institutions (strongly encouraged).
References:
- 3 letters of academic recommendation emailed directly from the letter writer.
Send all questions and materials to datalab@ucdavis.edu with the subject line “DataLab Postdoc Application.” Applications will be reviewed as they are received. A Fall 2020 start date is expected. The position is open until filled.