Deep, Complex, Invertible Networks for Inversion of Transmission Effects in Multimode Optical Fibres
- Submitting institution
-
University of Glasgow
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11-02581
- Type
- E - Conference contribution
- DOI
-
-
- Title of conference / published proceedings
- 32nd Conference on Neural Information Processing Systems (NeurIPS 2018)
- First page
- 1
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- December
- Year of publication
- 2018
- URL
-
http://eprints.gla.ac.uk/176148/
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
3
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- ORIGINALITY: First demonstration of general imaging, correcting distortions introduced in long, curving optical multimode-fibres, without phase measurements (making the sensing much simpler - significant breakthrough on physics side). Enabled by interdisciplinary development and implementation of a novel, complex-valued, invertible neural network with regularisers inspired by physical constraints. SIGNIFICANCE: Impact potential in medical imaging (ultrafine fibres cause less damage), communications using multimode-fibre’s increased capacity cf single-mode. RIGOUR: Theoretical consistency with physical constraints. Implemented, with empirical effectiveness demonstrated on novel experimental data. Published at top ML conference, and later, coupled physics-facing paper in Nature Communications (https://www.nature.com/articles/s41467-019-10057-8) is highly-cited.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -