Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Learning Robust Visual Place Recognition Combining Appearance, Sequence and Structure
Build a Neural Network that takes as input an image, a sequence of previous images, and the geometry extracted from that, and produces as output a descriptor which can be used for place recognition.
Keywords: CNN deep learning place recognition loop closure SLAM robotics computer vision VO
Recognizing a previously visited place is challenging when the place is subject to changes in appearance due to for example wheather, time-of-day or seasonal changes. Many previous works tackle the problem based on appearance, sequence or structure.
Recognizing a previously visited place is challenging when the place is subject to changes in appearance due to for example wheather, time-of-day or seasonal changes. Many previous works tackle the problem based on appearance, sequence or structure.
In this project, you will attempt to unify these approaches, while using machine learning techniques to your advantage. You will evaluate your work on a publicly available dataset that has been recorded on the same trajectory across the span of a year.
In this project, you will attempt to unify these approaches, while using machine learning techniques to your advantage. You will evaluate your work on a publicly available dataset that has been recorded on the same trajectory across the span of a year.
Titus Cieslewski ( titus at ifi.uzh.ch ), APPLY VIA EMAIL, ATTACH CV AND TRANSCRIPT! Required skills: Linux, Python, ability to read C++ code. Desirable skill: Tensorflow or similar.
Titus Cieslewski ( titus at ifi.uzh.ch ), APPLY VIA EMAIL, ATTACH CV AND TRANSCRIPT! Required skills: Linux, Python, ability to read C++ code. Desirable skill: Tensorflow or similar.