Is there "fat" (overheads) in cloud computing infrastructure software that can be trimmed? Would doing so help ameliorate the need for frequent hardware refreshes and extend the life of existing hardware? In this paper, we demonstrate that, indeed, there is "fat" that can be trimmed by using specialized OS-based software stacks. Doing so can allow decade-old computers to be used for critical cloud infrastructure services, potentially yielding 3x improvements in efficiency compared to standard software stacks on newer hardware. The implications of these results raise the possibility of exploiting OS optimizations to reduce server hardware obsolescence. Further, it suggests the importance of addressing the key portability challenges of specialized OS stacks.

This paper presents a system-level method for achieving the rapid deployment and high-density caching of serverless functions in a FaaS environment. For reduced start times, functions are deployed from unikernel snapshots, bypassing expensive initialization steps. To reduce the memory footprint of snapshots we apply page-level sharing across the entire software stack that is required to run a function. We demonstrate the effects of our techniques by replacing Linux on the compute node of a FaaS platform architecture. With our prototype OS, the deployment time of a function drops from 100s of milliseconds to under 10 ms. Platform throughput improves by 51x on workload composed entirely of new functions. We are able to cache over 50,000 function instances in memory as opposed to 3,000 using standard OS techniques. In combination, these improvements give the FaaS platform a new ability to handle large-scale bursts of requests.

Recent research has shown that generative adversarial networks (GANs) have been successfully applied in aligning features for domain adaptation. However, the extracted features might lose category-specific information, because they are distinguished as either a source or a target during the adversarial training. To address this issue, a two-stage training framework consisting of two sets of GANs and a dedicated classifier is proposed in this study. In the pretraining stage, we use an encoder-decoder-classifier structure to obtain discriminative and representative features of the source domain and use it as a reference in the subsequent training. In the adversarial training stage, two sets of GANs are used to align target-domain features with those of the source domain and transfer samples of the target domain to the source domain simultaneously. A dedicated classifier is trained along with the adversarial loss to force the generated features of the target domain to carry category-specific information, which significantly improves classification performance. The features of the source domain stay intact in the adversarial training stage. Thus, our approach can alleviate the training burden of GANs. The proposed method has been validated on digital datasets and office31 datasets. Experimental results demonstrate that an average accuracy of 96.5% and 95.1% is achieved, which leads to superior or comparable performance to state-of-the-art results.

Fast assessment of damaged buildings is important for post-disaster rescue operations. Building damage detection leveraging image processing and machine learning techniques has become a popular research focus in recent years. Although supervised learning approaches have made considerable improvement for damaged building assessment, rapidly deploying supervised classification is still difficult due to the complexity in obtaining a large number of labeled samples in the aftermath of disasters. We propose an unsupervised self-attention domain adaptation (USADA) model, which transforms instances of the source domain to those of the target domain in pixel space, to address the aforementioned issue. The proposed USADA consists of three parts: a set of generative adversarial networks (GANs), a classifier, and a self-attention module. The GAN adapts source domain images to ensure their similarity to target domain images. Once adapted, the classifier can be trained using the adapted images along with the original images of the source domain to classify damaged buildings. The self-attention module is introduced to maintain the foreground of the generated images conditioned on source domain images for generating plausible samples. As a case study, aerial images of Hurricane Sandy, Maria, and Irma, are used as the source and target domain datasets in our experiments. Experimental results demonstrate that classification accuracies of 68.1% and 84.1% are achieved, and our method obtains improvements of 2.0% and 3.6% against pixel-level domain adaptation, which is the basis of our model.

Building assessment is highly prioritized during rescue operations and damage relief after hurricane disasters. Although machine learning has made remarkable improvement in building damage classification, it remains challenging because classifiers must be trained using a massive amount of labeled data. Furthermore, data labeling is labor intensive, costly, and unavailable after a disaster. To address this issue, we propose an unsupervised domain adaptation method with aligned discriminative and representative features (ADRF), which leverage a substantial amount of labeled data of relevant disaster scenes for new classification tasks. The remote sensing imageries of different disasters are collected using different sensors, viewpoints, times, even at various places. Compared with the public datasets used in the domain adaptation community, the remote sensing imageries are more complicated which exhibit characteristics of lower discrimination between categories and higher diversity within categories. As a result, pursuing domain invariance is a huge challenge. To achieve this goal, we build a framework with ADRF to improve the discriminative and representative capability of the extracted features to facilitate the classification task. The ADRF framework consists of three pipelines: a classifier for the labeled data of the source domain and one autoencoder each for the source and target domains. The latent variables of autoencoders are forced to observe unit Gaussian distributions by minimizing the maximum mean discrepancy (MMD), whereas the marginal distributions of both domains are aligned via the MMD. As a case study, two challenging transfer tasks using the hurricane Sandy, Maria, and Irma datasets are investigated. Experimental results demonstrate that ADRF achieves overall accuracy of 71.6% and 84.1% in the transfer tasks from dataset Sandy to dataset Maria and dataset Irma, respectively.

Hyperspectral image classification has been proved significant in remote sensing field. Traditional classification methods have meet bottlenecks due to the lack of remote sensing background knowledge or high dimensionality. Deep learning based methods, such as deep convolutional neural network (CNN), can effectively extract high level features from raw data. But the training of deep CNN is rather time-consuming. The general purpose graphic processing units (GPUs) have been considered as one of the most common co-processors that can help accelerate deep learning applications. In this paper we propose a GPU-based Cube CNN (GCN) framework for hyperspectral image classification. First, a Parallel Neighbor Pixels Extraction (PNPE) algorithm is designed to enable the framework directly loading raw hyperspectral image into GPU's global memory, and extracting samples into data cube. Then, based on the peculiarity of hyperspectral image and cube convolution, we propose a novel Cube CNN-to-GPU mapping mechanism that transfers the training of Cube CNN to GPU effectively. Finally, the mini-batch gradient descent(MBGD) algorithm is improved with Computing United Device Architecture(CUDA) multi-streaming technique, which further speeds up network training in GCN framework. Experiments on KSC dataset, PU dataset and SA dataset show that, compared with state-of-art framework Caffe, we achieve up to 83% and 67% reduction in network training time without losing accuracy, when using SGD (Stochastic Gradient Descent) and MBGD algorithm respectively. Experiments across different GPUs show the same performance trend, which demonstrates the good extendibility of GCN framework.

CNNs (convolutional neural networks) have been proved to be efficient deep learning models that can directly extract high level features from raw data. In this paper, a novel CCS (Cube-CNN-SVM) method is proposed for hyperspectral image classification, which is a spectral-spatial feature based hybrid model of CNN and SVM (support vector machine). Different from most of traditional methods that only take spectral information into consideration, a target pixel and the spectral information of its neighbors are organized into a spectral-spatial multi-feature cube used in hyperspectral image classification. It is a straightforward but valid spatial strategy that can easily improve classification accuracy without extra modification of deep CNN's structure except the size of input layer and convolutional kernel. Our deep CNN consists of the input layer, convolutional layer, max pooling layer, full connection layer and output layer. To further improve hyperspectral image classification accuracy, SVM is trained as hyperspectral image classifier with the features extracted by deep CNN from spectral-spatial fusion information. Three hyperspectral image datasets such as the KSC (Kennedy Space Center), PU (Pavia University Scene) and Indian Pines are used to evaluate the performance of CCS method. Experimental results indicate that the hyperspectral image classification can be improved efficiently with the spectral-spatial fusion strategy and CCS method. Firstly, it is easy to implement the spatial strategy to improve classification accuracy about 4% compared with only spectral information used for classification, in which 98.49% is gained on the KSC dataset. Secondly, CCS method can further improve classification accuracy about 1%~3% compared to the best performance of deep CNN, in which 99.45% is gained on the PU dataset.

The wide use of micro bloggers such as Twitter offers a valuable and reliable source of information during natural disasters. The big volume of Twitter data calls for a scalable data management system whereas the semi-structured data analysis requires full-text searching function. As a result, it becomes challenging yet essential for disaster response agencies to take full advantage of social media data for decision making in a near-real-time fashion. In this work, we use Lucene to empower HBase with full-text searching ability to build a scalable social media data analytics system for observing and analyzing human behaviors during the Hurricane Sandy disaster. Experiments show the scalability and efficiency of the system. Furthermore, the discovery of communities has the benefit of identifying influential users and tracking the topical changes as the disaster unfolds. We develop a novel approach to discover communities in Twitter by applying spectral clustering algorithm to retweet graph. The topics and influential users of each community are also analyzed and demonstrated using Latent Semantic Indexing (LSI).

The knowledge on the internet is uncertain, quickly updated and from multi-sources. The information we extract from a field in webpages is usually one-sided and contains incorrect data. This paper proposed an algorithm to extract data from internet web pages or text documents by using the association strength. The algorithm combines knowledge extraction method of the text mining with the technology of the intelligence analysis to the data. It uses ontology theory to describe the knowledge, and automatically extract the knowledge from the web pages or text documents which is returned by the search engine.