Use this URL to cite or link to this record in EThOS:
Title: User assisted computer vision for biology
Author: Rennert, Peter
ISNI:       0000 0004 7964 8757
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Restricted access.
Access from Institution:
Computer vision is a broad field that combines aspects of image and signal processing with machine learning to solve tasks that involve analysis of image and video data. Biology is a science that increasingly produces such data on a scale that requires automation to analyse it. Challenges in biology include the high variability of the data, the unpredictability of the data and the novelty of computer vision and machine learning to the field. In this thesis we identify the lack of labelling tools which support exploration, annotation and visualisation of scientific recordings over a long time-span as the main bottleneck for computer vision being used in biology. As a solution to that problem, we present two new tools, AudioTagger and VideoTagger, which were used in large scale studies. We present the results of a 3 month life-span behavioural assay which was conducted using VideoTagger. We apply computer vision to count D. melanogaster eggs and introduce an interface which supports iterative learning and annotation to train the classifier. We show promising results for user assisted computer vision on the example of a user guided motion prediction system.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available