Search

Foundation models trained on large-scale dataset gain a recent surge in CV and NLP. In contrast, development in biomedical domain lags far behind due to data scarcity. To address this issue, we build and release PMC-OA, a biomedical dataset with 1.6M image-caption pairs collected from PubMedCentral's OpenAccess subset, which is 8 times larger than before. PMC-OA covers diverse modalities or diseases, with majority of the image-caption samples aligned at finer-grained level, i.e., subfigure and subcaption. While pretraining a CLIP-style model on PMC-OA, our model named PMC-CLIP achieves state-of-the-art results on various downstream tasks, including image-text retrieval on ROCO, MedMNIST image classification, Medical VQA, i.e. +8.1% R@10 on image-text retrieval, +3.9% accuracy on image classification., Comment: 10 pages, 3 figures

The continuous rapid development of new infrastructure load represented by electric vehicles (EVs) has brought new opportunities and challenges to the power system, as well as new propositions for traditional power system load modelling. It is of great practical significance to study the planning and operation of power systems considering EVs and other new infrastructure loads. Based on the analysis of the real historical data of EVs, this paper proposes an EV load modelling method based on the charging power scenario model. Based on the key variables of EV charging, the proposed model considers the joint distribution model of the uncertainty and correlation of the key variables of EV charging. Power scenarios are aggregated to obtain the EV load curve. Finally, the actual EV charging power data is used to verify the effectiveness of the proposed method.

Enhancement of transmitting power is an effective way to improve the depth of electrical prospecting. The objective of this method is to provide adequate offset observation in order to minimize source field effect and to reach a deeper depth of exploration. In addition, placement of electrically excited field source underground close to the exploration object and underground is a new prospecting technology in metal exploration. This method is implemented using existing pit, mine or borehole with the purpose of obtaining stronger and improved information on both the exploration depth and the definition of the anomalous body. Therefore, we have developed a transmitter that can be used as the excitation device for artificial electromagnetic source. It uses a type of large power inverse rectangular wave that works in an automatic sweep manner, the frequency of the transmitter signal range between 10 and 0.01 Hz, the maximum transmitter power is 48 KW, the maximum transmitter current is 60 A, it can realize the on-time planned power supply transmitting with the GPS clock and real time clock (RTC), and also record the information of several auxiliary status. We have been able to certify through experiments that the performance of the device is reliable and stable. And that the device can satisfy the demand for electromagnetic prospecting of metal mining, prospecting for deep and concealed ore bodies and the exploration of sustainable resources for solving resource mining crisis.