MASTER`S THESIS: Deep Learning Point Detection in Images

Deep Learning has revolutionized object detection in images. Most commonly objects are detected by drawing a bounding box around them, e.g., using Faster-R-CNN or Yolo deep neural network architectures. For industrial applications detection of points instead of bounding boxes is in many cases more useful. There is scientific literature on the point detection also, including:

Duan et al. “CenterNet: Keypoint Triplets for Object Detection”

Ribera et al. “Locating Objects Without Bounding Boxes”

The task of this thesis topic is to evaluate and possibly improve the point detection methods for industrial machine vision applications, such as autonomous bin picking and autonomous ground vehicles.

Prerequisites:

• Machine Learning & Computer Vision knowledge
• Programming skills

Contact:

For more information, contact R&D manager Ola Friman (ola.friman@sick.se) and apply below at "ANSÖKAN"  before  the 19th of October.  Please notice: Application must include CV, Cover letter and LADOK (study results including courses).