My main research interest is in computer vision, specifically semantic segmentation for scene understanding and end to end learning for steering angle regression.

Semantic Segmentation

Semantic segmentation is the task of taking an image and assigning a class to each pixel. An example is shown below.

Original Image Original Image

Semantically Segmented Image Segmented Image

State of the art algorithms rely on Convolutional Neural Networks (CNN).

this section is under construction

Publications

  1. Towards Pixel-Level OOD Detection for Semantic Segmentation
    Angus, M.;
    2019, Thesis.
    https://uwspace.uwaterloo.ca/items/593a5994-cb8a-4f99-8006-fa95496a1407
  2. Efficacy of pixel-level OOD detection for semantic segmentation
    Angus, M.; Czarnecki, K.; Salay, R.;
    2018, Preprint.
    https://arxiv.org/abs/1911.02897
  3. Unlimited Road-scene Synthetic Annotation (URSA) Dataset
    Angus, M.; ElBalkini, M.; Khan, S.; Harakeh, A.; Andrienko, O.; Reading, C.; Waslander, S.; Czarnecki, K.;
    IEEE 21st ITSC 2018, Accepted.
    https://arxiv.org/abs/1807.06056
  4. Trajectory Prediction of Traffic Agents at Urban Intersections Through Learned Interactions
    Sarkar, A.; Czarnecki, K.; Angus, M.; Li, C.; Waslander, S.;
    IEEE 20th ITSC 2017, Accepted.
    https://ieeexplore.ieee.org/document/8317731/