• Automated Speech Generation from UN General Assembly Statements: Mapping Risks in AI Generated Texts

    J. Bullock, M. Luengo-Oroz

    Published in AI for Social Good Workshop, 36th International Conference on Machine Learning (ICML)


    Automated text generation has been applied broadly in many domains such as marketing and robotics, and used to create chatbots, product reviews and write poetry. The ability to synthesize text, however, presents many potential risks, while access to the technology required to build generative models is becoming increasingly easy. This work is aligned with the efforts of the United Nations and other civil society organisations to highlight potential political and societal risks arising through the malicious use of text generation software, and their potential impact on human rights. As a case study, we present the findings of an experiment to generate remarks in the style of political leaders by fine-tuning a pretrained recurrent neural network language model on a dataset of speeches made at the UN General Assembly. This work highlights the ease with which this can be accomplished, as well as the threats of combining these techniques with other technologies.

  • XNet: a convolutional neural network (CNN) implementation for medical x-ray image segmentation suitable for small datasets

    J. Bullock, C. Cuesta-Lázaro, A. Quera-Bofarull

    Published in Proc. SPIE 10953, Medical Imaging 2019: Biomedical Applications in Molecular, Structural, and Functional Imaging (2019)


    X-Ray image enhancement, along with many other medical image processing applications, requires the segmentation of images into bone, soft tissue, and open beam regions. We apply a machine learning approach to this problem, presenting an end-to-end solution which results in robust and efficient inference. Since medical institutions frequently do not have the resources to process and label the large quantity of X-Ray images usually needed for neural network training, we design an end-to-end solution for small datasets, while achieving state-of-the-art results. Our implementation produces an overall accuracy of 92%, F1 score of 0.92, and an AUC of 0.98, surpassing classical image processing techniques, such as clustering and entropy-based methods while improving upon the output of existing neural networks used for segmentation in non-medical contexts.

  • Solving differential equations with neural networks: Applications to the calculation of cosmological phase transitions

    Maria Laura Piscopo, Michael Spannowsky, Philip Waite


    Starting from the observation that artificial neural networks are uniquely suited to solving optimization problems, and most physics problems can be cast as an optimization task, we introduce a novel way of finding a numerical solution to wide classes of differential equations. We find our approach to be very flexible and stable without relying on trial solutions, and applicable to ordinary, partial and coupled differential equations. We apply our method to the calculation of tunneling profiles for cosmological phase transitions, which is a problem of relevance for baryogenesis and stochastic gravitational wave spectra. Comparing our solutions with publicly available codes which use numerical methods optimized for the calculation of tunneling profiles, we find our approach to provide at least as accurate results as these dedicated differential equation solvers, and for some parameter choices, even more accurate and reliable solutions. In particular, we compare the neural network approach with two publicly available profile solvers, CosmoTransitions and BubbleProfiler, and give explicit examples where the neural network approach finds the correct solution while dedicated solvers do not. We point out that this approach of using artificial neural networks to solve equations is viable for any problem that can be cast into the form $\mathcal{F}(\vec{x}) = 0$, and is thus applicable to various other problems in perturbative and nonperturbative quantum field theory.

    Solving differential equations with neural networks: Applications to the calculation of cosmological phase transitions

  • Adversarially-trained autoencoders for robust unsupervised new physics searches

    Andrew Blance, Michael Spannowsky, Philip Waite

    Machine learning techniques in particle physics are most powerful when they are trained directly on data, to avoid sensitivity to theoretical uncertainties or an underlying bias on the expected signal. To be able to train on data in searches for new physics, anomaly detection methods are imperative, which can be realised by an autoencoder acting as an unsupervised classifier. The last source of uncertainties affecting the classifier are then experimental uncertainties in the reconstruction of the final-state objects. To mitigate their effect on the classifier and to allow for a realistic assessment of the method, we propose to combine the autoencoder with an adversarial neural network to remove its sensitivity to the smearing of the final-state objects. We quantify its effect and show that one can achieve a robust anomaly detection in resonance-induced top-antitop final states.”