Search for: All records

Corrosion is a prevalent issue in numerous industrial fields, causing expenses nearing $3 trillion or 4% of the GDP annually with safety threats and environmental pollution. To timely qualify and validate new corrosion-inhibiting materials on a large scale, accurate and efficient corrosion assessment is crucial. Yet it is hindered by a lack of automatic tools for expert-level corrosion segmentation of material science experimental images. Developing such tools is challenging due to limited domain-valid data, image artifacts visually similar to corrosion, various corrosion morphology, strong class imbalance, and millimeter-precision corrosion boundaries. To help the community address these challenges, we curate the first expert-level segmentation annotations for a real-world image dataset [1] for scientific corrosion segmentation. In addition, we design a deep learning-based model, called DeepSC-Edge that achieves guidance of ground-truth edge learning by adopting a novel loss that avoids over-fitting to edges. It also is enriched by integrating a class-balanced loss that improves segmentation with small area but crucial edges of interest for scientific corrosion assessment. Our dataset and methods pave the way to advanced deep-learning models for corrosion assessment and generation – promoting new research to connect computer vision and material science discovery. Once the appropriate approvals have been cleared, we expect to release the code and data at: https://arl.wpi.edu/ 

Deep learning often relies on the availability of a large amount of high-quality labeled data, which can be very limited in novel domains. To address such data scarcity, domain adaptation is one promising approach that allows for deep networks to leverage large amounts of available data from a source domain to enhance the model’s efficacy on the target domain of interest. However, while there is a plethora of alternate models for domain adaptation proposed over many years in the literature, there is a dearth of studies that objectively compare the relative effectiveness of these models in a rigorous, empirical study. To fill this gap, we provide a thorough, unbiased, empirical study of five state-of-the-art (SOTA) deep domain adaptation models proposed over the past 6 years whose codes are publicly available. Models are evaluated on the complex and diverse domain adaptation tasks featured in the DomainNet benchmark dataset as well as the popular Office-31 dataset. Our results suggest that (1) all 5 models perform similarly, on average, and do not even significantly beat the oldest model, and (2) counter to their intended purpose, the transfer loss functions in the literature do not contribute significantly to learning transferable representations. Our observations suggest that domain adaptation research needs to more thoroughly compare newly proposed models against existing works, along with assessing their loss functions’ utility thoroughly. Our code and data splits are made public for reproducibility of results by the community. 

Lightweight neural networks refer to deep networks with small numbers of parameters, which can be deployed in resource-limited hardware such as embedded systems. To learn such lightweight networks effectively and efficiently, in this paper we propose a novel convolutional layer, namely Channel-Split Recurrent Convolution (CSR-Conv), where we split the output channels to generate data sequences with length T as the input to the recurrent layers with shared weights. As a consequence, we can construct lightweight convolutional networks by simply replacing (some) linear convolutional layers with CSR-Conv layers. We prove that under mild conditions the model size decreases with the rate of O( 1 ). Empirically we demonstrate the state-of-the-art T2 performance using VGG-16, ResNet-50, ResNet-56, ResNet- 110, DenseNet-40, MobileNet, and EfficientNet as backbone networks on CIFAR-10 and ImageNet. Codes can be found on https://github.com/tuaxon/CSR Conv. 

Cross-modal vehicle localization is an important task for automated driving systems. This research proposes a novel approach based on LiDAR point clouds and OpenStreetMaps (OSM) via a constrained particle filter, which significantly improves the vehicle localization accuracy. The OSM modality provides not only a platform to generate simulated point cloud images, but also geometrical constraints (e.g., roads) to improve the particle filter’s final result. The proposed approach is deterministic without any learning component or need for labelled data. Evaluated by using the KITTI dataset, it achieves accurate vehicle pose tracking with a position error of less than 3 m when considering the mean error across all the sequences. This method shows state-of-the-art accuracy when compared with the existing methods based on OSM or satellite maps.