We propose a novel method for representing oriented objects in aerial images named Adaptive Period Embedding (APE). While traditional object detection methods represent object with horizontal bounding boxes, the objects in aerial images are oritented. Calculating the angle of object is an yet challenging task. While almost all previous object detectors for aerial images directly regress the angle of objects, they use complex rules to calculate the angle, and their performance is limited by the rule design. In contrast, our method is based on the angular periodicity of oriented objects. The angle is represented by two two-dimensional periodic vectors whose periods are different, the vector is continuous as shape changes. The label generation rule is more simple and reasonable compared with previous methods. The proposed method is general and can be applied to other oriented detector. Besides, we propose a novel IoU calculation method for long objects named length independent IoU (LIIoU). We intercept part of the long side of the target box to get the maximum IoU between the proposed box and the intercepted target box. Thereby, some long boxes will have corresponding positive samples. Our method reaches the 1 st place of DOAI2019 competition task1 (oriented object) held in workshop on Detecting Objects in Aerial Images in conjunction with IEEE CVPR 2019.

We study a new problem setting of information extraction (IE), referred to as text-to-table . In text-to-table, given a text, one creates a table or several tables expressing the main content of the text, while the model is learned from text-table pair data. The problem setting differs from those of the existing methods for IE. First, the extraction can be carried out from long texts to large tables with complex structures. Second, the extraction is entirely data-driven, and there is no need to explicitly define the schemas. As far as we know, there has been no previous work that studies the problem. In this work, we formalize textto-table as a sequence-to-sequence (seq2seq) problem. We first employ a seq2seq model finetuned from a pre-trained language model to perform the task. We also develop a new method within the seq2seq approach, exploiting two additional techniques in table generation: table constraint and table relation embeddings.We consider text-to-table as an inverse problem of the well-studied table-to-text, and make use of four existing table-to-text datasets in our experiments on text-to-table. Experimental results show that the vanilla seq2seq model can outperform the baseline methods of using relation extraction and named entity extraction. The results also show that our method can further boost the performances of the vanilla seq2seq model. We further discuss the main challenges of the proposed task. The code and data are available at https://github. com/shirley-wu/text_to_table. 1

No abstract

No abstract

No abstract

Internet data centers are growing rapidly in recent years and they operate with intensive energy activity. Combined cooling, heating and power (CCHP) brings new opportunities for reducing the electricity cost in internet data centers. The main objective of this study is to optimize the energy resources scheduling in the data center coupled energy nets considering the involvement of CCHP and different demand response techniques. In this paper, internet data center coupled energy nets are proposed, where power grid, solar photovoltaic, CCHP, and battery energy storage systems are the primary energy sources. The adjunct residential buildings and commercial buildings near the internet data centers are also included in the proposed energy nets, where different types of load and demand response characteristics are utilized. A two-stage optimized energy management model considering the coordinated operation of CCHP and demand response technologies is established for internet data center coupled energy nets. In the day-ahead stage, the control objective is to minimize system cost while satisfying various constraints. Consider the electricity tariff chance between day-ahead market and real-time market, real-time control is implemented to minimize the imbalance cost between two electricity markets. Case studies are conducted on a practical internet data center coupled energy nets in Foshan City, China. It is observed that the proposed control framework can optimally schedule the energy resources in the energy network to meet system demand and improve the energy efficiency. The economic evaluation demonstrates that the proposed control scheme reduces system daily cost by 22.01%.INDEX TERMS Combined cooling, heating and power, data centers, mixed integer linear programming, renewable energy sources, two-stage optimal scheduling.

Data augmentation, which refers to manipulating the inputs (e.g., adding random noise, masking specific parts) to enlarge the dataset, has been widely adopted in machine learning. Most data augmentation techniques operate on a single input, which limits the diversity of the training corpus. In this paper, we propose a simple yet effective data augmentation technique for neural machine translation, mixSeq, which operates on multiple inputs and their corresponding targets. Specifically, we randomly select two input sequences, concatenate them together as a longer input as well as their corresponding target sequences as an enlarged target, and train models on the augmented dataset. Experiments on nine machine translation tasks demonstrate that such a simple method boosts the baselines by a nontrivial margin. Our method can be further combined with single-input based data augmentation methods to obtain further improvements.

For microgrids (MGs) optimal operation, one heated topic is the uncertainty management associated with renewable variations and electricity load forecasting errors. On the other hand, the networking of MGs is receiving an increasing attention in recent years. In this paper, an interactive energy management strategy is developed for high renewable-penetrated MGs. The control method includes two steps. In the first step, a local optimization is proposed for each microgrid to minimize the operation cost during the whole scheduling periods. In the second step, a global optimization is conducted for networked microgrids. CVaR based risk averse measure is introduced here to provide a risk-hedging strategy for microgrids energy management. Formulated models are solved by the easily implemented and computationally inexpensive mix integer linear programming (MILP) solver. Case studies demonstrate the feasibility of the proposed method by identifying optimal scheduling results.