More Like this

1. PUB-SalNet: A Pre-Trained Unsupervised Self-Aware Backpropagation Network for Biomedical Salient Segmentation

   https://doi.org/10.3390/a13050126  

   Chen, Feiyang; Jiang, Ying; Zeng, Xiangrui; Zhang, Jing; Gao, Xin; Xu, Min (May 2020, Algorithms)

   null (Ed.)

   Salient segmentation is a critical step in biomedical image analysis, aiming to cut out regions that are most interesting to humans. Recently, supervised methods have achieved promising results in biomedical areas, but they depend on annotated training data sets, which requires labor and proficiency in related background knowledge. In contrast, unsupervised learning makes data-driven decisions by obtaining insights directly from the data themselves. In this paper, we propose a completely unsupervised self-aware network based on pre-training and attentional backpropagation for biomedical salient segmentation, named as PUB-SalNet. Firstly, we aggregate a new biomedical data set from several simulated Cellular Electron Cryo-Tomography (CECT) data sets featuring rich salient objects, different SNR settings, and various resolutions, which is called SalSeg-CECT. Based on the SalSeg-CECT data set, we then pre-train a model specially designed for biomedical tasks as a backbone module to initialize network parameters. Next, we present a U-SalNet network to learn to selectively attend to salient objects. It includes two types of attention modules to facilitate learning saliency through global contrast and local similarity. Lastly, we jointly refine the salient regions together with feature representations from U-SalNet, with the parameters updated by self-aware attentional backpropagation. We apply PUB-SalNet for analysis of 2D simulated and real images and achieve state-of-the-art performance on simulated biomedical data sets. Furthermore, our proposed PUB-SalNet can be easily extended to 3D images. The experimental results on the 2d and 3d data sets also demonstrate the generalization ability and robustness of our method. 

   more » « less
2. SUMDocS: Surrounding-aware Unsupervised Multi-Document Summarization

   https://doi.org/10.1137/1.9781611976700.54  

   Qi Zhu, Fang Guo (April 2021, Proceedings of the 2021 {SIAM} International Conference on Data Mining, {SDM} 2021)

   Carlotta Demeniconi, Ian Davidson: (Ed.)

   Multi-document summarization, which summarizes a set of documents with a small number of phrases or sentences, provides a concise and critical essence of the documents. Existing multi-document summarization methods ignore the fact that there often exist many relevant documents that provide surrounding background knowledge, which can help generate a salient and discriminative summary for a given set of documents. In this paper, we propose a novel method, SUMDocS (Surrounding-aware Unsupervised Multi-Document Summarization), which incorporates rich surrounding (topically related) documents to help improve the quality of extractive summarization without human supervision. Speci fically, we propose a joint optimization algorithm to unify global novelty (i.e., category-level frequent and discriminative), local consistency (i.e., locally frequent, co-occurring), and local saliency (i.e., salient from its surroundings) such that the obtained summary captures the characteristics of the target documents. Extensive experiments on news and scientifi c domains demonstrate the superior performance of our method when the unlabeled surrounding corpus is utilized. 

   more » « less
3. Adaptive Compression-Aware Split Learning and Inference for Enhanced Network Efficiency

   https://doi.org/10.1145/3687471  

   Mudvari, Akrit; Vainio, Antero; Ofeidis, Iason; Tarkoma, Sasu; Tassiulas, Leandros (November 2024, ACM Transactions on Internet Technology)

   The growing number of AI-driven applications in mobile devices has led to solutions that integrate deep learning models with the available edge-cloud resources. Due to multiple benefits such as reduction in on-device energy consumption, improved latency, improved network usage, and certain privacy improvements, split learning, where deep learning models are split away from the mobile device and computed in a distributed manner, has become an extensively explored topic. Incorporating compression-aware methods (where learning adapts to compression level of the communicated data) has made split learning even more advantageous. This method could even offer a viable alternative to traditional methods, such as federated learning techniques. In this work, we develop an adaptive compression-aware split learning method (“deprune”) to improve and train deep learning models so that they are much more network-efficient, which would make them ideal to deploy in weaker devices with the help of edge-cloud resources. This method is also extended (“prune”) to very quickly train deep learning models through a transfer learning approach, which tradesoff little accuracy for much more network-efficient inference abilities. We show that the “deprune” method can reduce network usage by 4× when compared with a split-learning approach (that does not use our method) without loss of accuracy, while also improving accuracy over compression-aware split-learning by up to 4 percent. Lastly, we show that the “prune” method can reduce the training time for certain models by up to 6× without affecting the accuracy when compared against a compression-aware split-learning approach. 

   more » « less
4. Input-output Data-driven Modeling and MIMO Predictive Control of an RCCI Engine Combustion

   Khoshbakht Irdmousa, Behrouz; Naber, Jeffrey Donald; Mohammadpour Velni, Javad; Borhan, Hoseinali; Shahbakhti, Mahdi (January 2021, IFAC Modeling, Estimation and Control Conference (MECC))

   null (Ed.)

   This study presents a data-driven identi cation method based on Kernelized Canonical Correlation Analysis (KCCA) approach to generate a state-space Linear Parameter-Varying (LPV) dynamic representation for the RCCI engine combustion. An LPV model is used to estimate RCCI combustion phasing (CA50) and indicated mean eective pressure (IMEP) based on fuel injection timing and quantity. The proposed data-driven method does not require prior knowledge of the plant model states and adjusts number of states to increase the accuracy of the identi ed state-space model. The results demonstrate that the proposed data-driven KCCA-LPV approach provides a dependable technique to establish a fast and reasonably accurate RCCI combustion model. The established model is then incorporated in a design of a constrained MIMO Model Predictive Controller (MPC) to track desired crank angle for 50% fuel burnt and IMEP at various engine conditions. The controller performance results demonstrate that the established data-driven constrained MPC combustion controller can follow desired CA50 and IMEP with less than 1.5 CAD and 37 kPa error, respectively. 

   more » « less
5. Digital twins of the Earth: can they keep up?

   https://doi.org/10.3389/fsci.2024.1371049  

   Tague, Christina (March 2024, Frontiers in Science)

   • First-generation Earth system digital twins, such as the Digital Twin Earth (DTE) for hydrology, create important opportunities for “learning by doing” that will ensure DTEs evolve to provide credible, reliable, and useful information. • Recent DTEs for hydrology demonstrate the complexity of the cyberinfrastructure needed to support the integration of a diversity of high-resolution datasets—often through machine learning techniques— while also providing initial insights into how critical errors in these approaches might be identi!ed. • To remain useful, DTEs will need to be able to continuously evolve—this will require innovations in visualization, cross-disciplinary collaboration, and complementary tools that draw from advances in relevant research communities. 

   more » « less