Models, code, and papers for "Geoff French":
Consistency regularization describes a class of approaches that have yielded ground breaking results in semi-supervised classification problems. Prior work has established the cluster assumption -- under which the data distribution consists of uniform class clusters of samples separated by low density regions -- as key to its success. We analyse the problem of semantic segmentation and find that the data distribution does not exhibit low density regions separating classes and offer this as an explanation for why semi-supervised segmentation is a challenging problem. We adapt the recently proposed CutMix regularizer for semantic segmentation and find that it is able to overcome this obstacle, leading to a successful application of consistency regularization to semi-supervised semantic segmentation.
This paper explores the use of self-ensembling for visual domain adaptation problems. Our technique is derived from the mean teacher variant (Tarvainen et al., 2017) of temporal ensembling (Laine et al;, 2017), a technique that achieved state of the art results in the area of semi-supervised learning. We introduce a number of modifications to their approach for challenging domain adaptation scenarios and evaluate its effectiveness. Our approach achieves state of the art results in a variety of benchmarks, including our winning entry in the VISDA-2017 visual domain adaptation challenge. In small image benchmarks, our algorithm not only outperforms prior art, but can also achieve accuracy that is close to that of a classifier trained in a supervised fashion.
In this paper we test the use of a deep learning approach to automatically count Wandering Albatrosses in Very High Resolution (VHR) satellite imagery. We use a dataset of manually labelled imagery provided by the British Antarctic Survey to train and develop our methods. We employ a U-Net architecture, designed for image segmentation, to simultaneously classify and localise potential albatrosses. We aid training with the use of the Focal Loss criterion, to deal with extreme class imbalance in the dataset. Initial results achieve peak precision and recall values of approximately 80%. Finally we assess the model's performance in relation to inter-observer variation, by comparing errors against an image labelled by multiple observers. We conclude model accuracy falls within the range of human counters. We hope that the methods will streamline the analysis of VHR satellite images, enabling more frequent monitoring of a species which is of high conservation concern.
Mental illness affects a significant portion of the worldwide population. Online mental health forums can provide a supportive environment for those afflicted and also generate a large amount of data which can be mined to predict mental health states using machine learning methods. We benchmark multiple methods of text feature representation for social media posts and compare their downstream use with automated machine learning (AutoML) tools to triage content for moderator attention. We used 1588 labeled posts from the CLPsych 2017 shared task collected from the Reachout.com forum (Milne et al., 2019). Posts were represented using lexicon based tools including VADER, Empath, LIWC and also used pre-trained artificial neural network models including DeepMoji, Universal Sentence Encoder, and GPT-1. We used TPOT and auto-sklearn as AutoML tools to generate classifiers to triage the posts. The top-performing system used features derived from the GPT-1 model, which was finetuned on over 150,000 unlabeled posts from Reachout.com. Our top system had a macro averaged F1 score of 0.572, providing a new state-of-the-art result on the CLPsych 2017 task. This was achieved without additional information from meta-data or preceding posts. Error analyses revealed that this top system often misses expressions of hopelessness. We additionally present visualizations that aid understanding of the learned classifiers. We show that transfer learning is an effective strategy for predicting risk with relatively little labeled data. We note that finetuning of pretrained language models provides further gains when large amounts of unlabeled text is available.