We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Research & business

Research lectures, seminars and events

The events listed in this area are research seminars, workshops and lectures hosted by Durham University departments and research institutes. If you are not a member of the University, but  wish to enquire about attending one of the events please contact the organiser or host department.


November 2020
October 2020 December 2020
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30

Events for 13 November 2020

Marcin Lis: On delocalization in the six-vertex model.

1:00pm, Zoom

In this talk I will show that the six-vertex model with parameter c in [\sqrt 3, 2] on a square lattice torus has an ergodic infinite-volume limit as the size of the torus grows to infinity.
Moreover I will prove that for c in [\sqrt{2+\sqrt 2}, 2], the associated height function on the infinite square lattice has unbounded variance.

The proof relies on an extension of the Baxter--Kelland--Wu representation of the six-vertex model
to multi-point correlation functions of the associated spin model.
Other crucial ingredients are the uniqueness and percolation properties of the critical random cluster measure for $q\in[1,4]$, and recent results relating the decay of correlations
in the spin model with the delocalization of the height function.


Contact for more information about this event.

Neelanjan Bhowmik and Jack W Barker: The Good, the Bad and the Ugly: Evaluating Convolutional Neural Networks for Prohibited Item Detection Using Real and Synthetically Composited X-ray Imagery

1:00pm, Online

Detecting prohibited items in X-ray security imagery is pivotal in maintaining border and transport security against a wide range of threat profiles. Convolutional Neural Networks (CNN) with the support of a significant volume of data have brought advancement in such automated prohibited object detection and classification. However, collating such large volumes of X-ray security imagery remains a significant challenge. This work opens up the possibility of using synthetically composed imagery, avoiding the need to collate such large volumes of hand-annotated real-world imagery. Here we investigate the difference in detection performance achieved using real and synthetic X-ray training imagery for CNN architecture detecting three exemplar prohibited items, {Firearm, Firearm Parts, Knives}, within cluttered and complex X-ray security baggage imagery. Our extended evaluation demonstrates both challenge and promise of using synthetically composed images to diversify the X-ray security training imagery for automated detection algorithm training. Additionally, what's the future on synthetic data generation techniques: we introduce briefly our initial works on synthetically composited X-ray imagery using Generative Adversarial Networks (GAN).

Contact for more information about this event.

Federico Carta: Supersymmetry Enhancement


In the last couple of years it was discovered that some 4d N=1 quantum field theories flow in the IR to 4d N=2 superconformal field theories (often of generalized Argyres-Douglas type), therefore showing a phenomenon of Supersymmetry Enhancement at the IR fixed point. Such flows are extremely useful in order to learn features of the IR non-lagrangian theory, by using the UV formulation to compute RG flow protected quantities. However, up to date it is not completely clear why such flows exist, and how the SUSY Enhancement mechanism works in detail. Limiting ourself to one class of such flows, usually referred as Maruyoshi-Song flows, we show how it is possible to understand the enhancement phenomenon in a geometric way, by realizing this setup in F-Theory. We also discuss how to understand the enhancement as an hyperkahler structure restoration on the moduli space of solutions of the (generalized) Hitchin system associated to such theories. Finally, we discuss a very large systematic scan that we performed in order to find more example of such flows. Having found none, we conjecture that this particular method is exhausted: all the N=1 lagrangian theories flowing to Argyres-Douglas theory by Maruyoshi-Song RG-flows have been already discovered, and there exist no more.

Contact for more information about this event.