Machine Learning Internship
1. Project Description:
The Upper Midwest Environmental Sciences Center is developing machine learning algorithms and tools for the automatic detection, enumeration, classification, and annotation of seabirds and other marine wildlife from digital aerial imagery. Student assistance is required for machine learning algorithm development, data processing and management scripting, and web application development. Student work shall also include the use of high-performance computing resources. Outcomes of this project seek to improve the efficiency, standardization, and accuracy of aerial waterfowl population surveys to inform harvest and other regulatory decisions, as well as to provide baseline wildlife information for environmental assessments and impact analyses of potential wildlife exposure to offshore energy development project in the United States.
Student(s) shall gain professional experience working with a federal science agency while furthering their knowledge of machine learning, data management, web development, and high-performance computing.
2. Duties:
The student(s) shall be expected to:
- write scripts for data processing and management, including the integration of high-performance computing resources;
- contribute to designing, developing, and refining an online annotation tool that will be used by wildlife experts to create labeled training data from digital aerial imagery;
- clean and use training data (i.e., annotated wildlife targets) to train and evaluate machine learning models;
- collaborate with a cross functional development team spanning data science, mathematics, geospatial sciences, and biology; and
- utilize email, instant messaging, and video conferencing applications to communicate with other remote team members.
3. Required expertise/skills:
- Computer science and/or data science experience; programming skills are required.
- Be proficient with Microsoft Outlook, Teams and Office.
- Be self-motivated, detail oriented, and able to work in an efficient and self-directed manner.
- Ability to multi-task and work collaboratively in a team setting.
- Be a United States citizen or possess a Permanent Resident Card.
Desired qualifications:
- Previous experience with the following is preferred, but not required: Python, Django, PostgreSQL, React/Redux, Linux, and Git.
4. Hours and Duration of employment:
The anticipated start date for this position is on or about August 01, 2021, and the position shall continue into July 31, 2023, if funds are available. Start date can be adjusted to accommodate completion of course work and final examination schedules. During this period, the total number of hours worked shall not exceed 2080 hours.
5. Description of Working Conditions:
Work will be performed on a computer (in a remote setting while current maximum telework posture remains in effect). Extended periods of sitting and prolonged use of computers will be required. The student(s) will be responsible for providing their own work setting while remote work continues.
6. Compensation:
Compensation is commensurate with the level of education and experience as follows:
$ 14.48 per hour for 1 year of college.
$ 16.26 per hour for 2 years of college.
Students are paid only for hours worked (with no holiday or leave benefits). Students will earn 1 hour of sick leave for every 30 hours of work.
Student is responsible for all costs of transportation to and from the principal duty station location. The Government does not provide housing, meals or other living expenses while working at the principal duty station. If travel is required, transportation for official travel shall be provided to and from the destination site(s). Costs of meals and lodging while on travel status shall be provided by the USGS in the same manner as for Federal employees, in accordance with the Government Travel Regulations. The position is a contract position, and not a federal government position. Successful candidate(s) shall be required to register as a contractor.
7. Principal Duty Station:
U.S. Geological Survey Upper Midwest Environmental Sciences Center
2630 Fanta Reed Road
La Crosse, WI 54603
Work shall be performed remotely until the current maximum telework posture has ended.