Machine learning methods have found many applications in Raman spectroscopy, especially for the identification of chemical species. However, almost all of these methods require non-trivial preprocessing such as baseline correction and/or PCA as an essential step. Here we describe our unified solution for the identification of chemical species in which a convolutional neural network is trained to automatically identify substances according to their Raman spectrum without the need of ad-hoc preprocessing steps. We evaluated our approach using the RRUFF spectral database, comprising mineral sample data. Superior classification performance is demonstrated compared with other frequently used machine learning algorithms including the popular support vector machine.

The version in the Kent Academic Repository may differ from the final published version. Users are advised to check http://kar.kent.ac.uk for the status of the paper. Users should always cite the published version of record.

A facial composite system is described for use in criminal investigations which has distinct advantages over current methods. Unlike traditional feature based methods, our approach uses both local and global facial models, allowing a witness to evolve plausible, photo-realistic face images in an intuitive way. The basic method combines random sampling from a facial appearance model (AM) with an evolutionary algorithm (EA) to drive the search procedure to convergence. Three variants of the evolutionary algorithm have been explored and their performance measured using a computer simulation of a human witness (virtual witness). Further system functionality, provided by local appearance models and transformations of the appearance space which respectively allow both local features and semantic facial attributes to be manipulated, is presented. Preliminary examples of composites generated with our system are presented which demonstrate the potential superiority of the evolutionary approach to composite generation.

Convolutional neural networks (CNN) have been shown to provide a good solution for classification problems that utilize data obtained from vibrational spectroscopy. Moreover, CNNs are capable of identification from noisy spectra without the need for additional preprocessing. However, their application in practical spectroscopy is limited due to two shortcomings. The e ectiveness of the classification using CNNs drops rapidly when only a small number of spectra per substance are available for training (which is a typical situation in real applications). Additionally, to accommodate new, previously unseen substance classes, the network must be retrained which is computationally intensive. Here we address these issues by reformulating a multi-class classification problem with a large number of classes, but a small number of samples per class, to a binary classification problem with su icient data available for representation learning. Namely, we define the learning task as identifying pairs of inputs as belonging to the same or di erent classes. We achieve this using a Siamese convolutional neural network. A novel sampling strategy is proposed to address the imbalance problem in training the Siamese Network. The trained network can e ectively classify samples of unseen substance classes using just a single reference sample (termed as one-shot learning in the machine learning community). Our results demonstrate be er accuracy than other practical systems to date, while allowing e ortless updates of the system's database with novel substance classes.

Student participant-witnesses produced 4 composites of unfamiliar faces with a system that uses a genetic algorithm to evolve appearance of artificial faces. Morphs of 4 composites produced by different witnesses (between-witness morphs) were judged better likenesses (Experiment 1) and were more frequently named (Experiment 2) by participants who were familiar with the target actors than were morphs of 4 composites produced by a single witness (within-witness morphs). Within-witness morphs were judged better likenesses and more frequently named than the best or the first-produced individual composites. The same results for likeness judgments were observed after possible artifacts in the comparison of between- and within-witness morphs were eliminated (Experiment 3). Experiment 4 showed that both internal and external features were better represented in morphs than in the original composites, although the representation of internal features improved more. The results suggest that morphing improves the representation of faces by reducing random error. Between-witness morphs yield more benefit than within-witness morphs by reducing consistent but idiosyncratic errors of individual witnesses. The experiments provide the first demonstration of an advantage for within-witness morphs produced using a single system. Experiment 2 provides the first demonstration of a reliable advantage for between-witness morphs in the most forensically relevant task: naming a composite of a familiar person produced by a witness who was unfamiliar with the target. Morphing would enhance the recognition of facial composites of criminals. Within-witness morphing provides a methodology for use in crimes in which the victim is the only witness.