Skip to main content

S4&5 Multidisciplinary Computer Science

Instructions​

There are two repositories associated with this course:

  • S4-ml (private repository), which contains an introduction to Python, then moves on to an overview of deep learning methods.
  • S4-imaging, which covers image storage and structure, and pixel-based approaches.

It is suggested that WildDrone students undertake the first intro notebook from S4-ml, then move to S4-imaging, before heading back to S4-ml to tackle the machine learning aspects.

Note that at-present S4-ml uses Google Colab as an online Python environment, whereas S4-imaging has been developed to run on a local Python environment. There are lots of ways to set Python up locally - here's one guide - but it's recommended that students follow whichever approach their institution recommends so that they can obtain local support.

Objectives​

To provide interdisciplinary education for multiple PhDs with different scientific backgrounds, via an e-learning course on programming and data science, image processing, and applied machine learning for computer vision purposes. The course will enable targeted education to teach undeveloped skills (e.g., programming and data science methods for ecologists and engineers).

Intended Learning Outcomes​

Students will be able to:

  1. Set up and use a Python programming environment, either online or locally, including scientific computing packages such as NumPy and SciPy.
  2. Understand computer imaging and interrogate images using pixel-based approaches.
  3. Use deep learning approaches for computer vision applications based on TensorFlow and/or PyTorch.

This course provides an introduction to Python programming, scientific computing, image processing, and machine learning. We will cover relevant python packages such as NumPy and SciPy, image-based approaches using OpenCV and we will also introduce deep learning approaches for computer vision applications based on TensorFlow and/or PyTorch. We will include theory and programming exercises.

Assessment​

  • Submission: Jupyter Notebook, graded pass/fail - details in due course.
  • Deadline: TBC