Pre-trained vision language models (VL) have seen a rise in recent years, achieving state-of-the-art performance on tasks such as visual question answering, image captioning, zero-shot tasks, text-to-image synthesis, and many more. However, these models are large, also requiring large amounts of data, which can hinder application on resource-limited settings. This paper proposes CAPIT (Cross Attention on Pre-trained Image and Text models), a novel architecture built on top of frozen pre-trained unimodal encoders, utilizing transferable knowledge from pre-trainined models through cross-attentional transformers. CAPIT is trained with a simple supervised task that learns to predict the correspondence between image-text pairs. It is then tested on zero-shot image classification. The proposed model contains less parameters and uses less data compared to other methods, while performing fairly well with trade-offs between compute requirements performance, showing promise for future work in resource-limited settings.

Feature pre-processing is an essential step for most machine learning algorithms. Pre-processing usually involves transforming the input data to a form that is more suitable for the learning algorithm. Various techniques can be used for pre-processing, such as feature scaling, feature extraction, and feature selection which help to improve the performance of the learning algorithm. This study develops novel pre-processing methods through a genetic programming approach. Genetic Algorithms were used to search for a combination of pre-processing operations that produced the best results of Multilayer Perceptrons on a set of binary classification datasets. The search findings show that these discovered methods, when combined with existing methods, statistically outperform existing methods by themselves on new datasets. Visualization of the effects its on synthetic data show that these discovered methods extend the range of the data and direct values away from the center of the data. This study provides practitioners with new methods that can be used as pre-processing techniques for machine learning algorithms.

In order to help development into analyzing the characteristics of adversarial sample generation in artificial neural networks, this work proposes a framework for an adversarial attack that utilizes neural image modification to generate an adversarial sample. This method proves to be effective in reducing a target network's accuracy in both untargeted and targeted attacks with good success rates. This method also shows some effectiveness against defensive distillation, but not transferrable between multiple models.