Tsinghua University

About: Tsinghua University is a education organization based out in Beijing, Beijing, China. It is known for research contribution in the topics: Computer science & Catalysis. The organization has 129978 authors who have published 200506 publications receiving 4549561 citations. The organization is also known as: Tsinghua & THU.

Recurrent convolutional neural network for object recognition

Abstract: In recent years, the convolutional neural network (CNN) has achieved great success in many computer vision tasks. Partially inspired by neuroscience, CNN shares many properties with the visual system of the brain. A prominent difference is that CNN is typically a feed-forward architecture while in the visual system recurrent connections are abundant. Inspired by this fact, we propose a recurrent CNN (RCNN) for object recognition by incorporating recurrent connections into each convolutional layer. Though the input is static, the activities of RCNN units evolve over time so that the activity of each unit is modulated by the activities of its neighboring units. This property enhances the ability of the model to integrate the context information, which is important for object recognition. Like other recurrent neural networks, unfolding the RCNN through time can result in an arbitrarily deep network with a fixed number of parameters. Furthermore, the unfolded network has multiple paths, which can facilitate the learning process. The model is tested on four benchmark object recognition datasets: CIFAR-10, CIFAR-100, MNIST and SVHN. With fewer trainable parameters, RCNN outperforms the state-of-the-art models on all of these datasets. Increasing the number of parameters leads to even better performance. These results demonstrate the advantage of the recurrent structure over purely feed-forward structure for object recognition.

Defect engineering of two-dimensional transition metal dichalcogenides

Abstract: Two-dimensional transition metal dichalcogenides (TMDs), an emerging family of layered materials, have provided researchers a fertile ground for harvesting fundamental science and emergent applications. TMDs can contain a number of different structural defects in their crystal lattices which significantly alter their physico-chemical properties. Having structural defects can be either detrimental or beneficial, depending on the targeted application. Therefore, a comprehensive understanding of structural defects is required. Here we review different defects in semiconducting TMDs by summarizing: (i) the dimensionalities and atomic structures of defects; (ii) the pathways to generating structural defects during and after synthesis and, (iii) the effects of having defects on the physico-chemical properties and applications of TMDs. Thus far, significant progress has been made, although we are probably still witnessing the tip of the iceberg. A better understanding and control of defects is important in order to move forward the field of Defect Engineering in TMDs. Finally, we also provide our perspective on the challenges and opportunities in this emerging field.

Rolling circle amplification: A versatile tool for chemical biology, materials science and medicine

Abstract: Rolling circle amplification (RCA) is an isothermal enzymatic process where a short DNA or RNA primer is amplified to form a long single stranded DNA or RNA using a circular DNA template and special DNA or RNA polymerases. The RCA product is a concatemer containing tens to hundreds of tandem repeats that are complementary to the circular template. The power, simplicity, and versatility of the DNA amplification technique have made it an attractive tool for biomedical research and nanobiotechnology. Traditionally, RCA has been used to develop sensitive diagnostic methods for a variety of targets including nucleic acids (DNA, RNA), small molecules, proteins, and cells. RCA has also attracted significant attention in the field of nanotechnology and nanobiotechnology. The RCA-produced long, single-stranded DNA with repeating units has been used as template for the periodic assembly of nanospecies. Moreover, since RCA products can be tailor-designed by manipulating the circular template, RCA has been employed to generate complex DNA nanostructures such as DNA origami, nanotubes, nanoribbons and DNA based metamaterials. These functional RCA based nanotechnologies have been utilized for biodetection, drug delivery, designing bioelectronic circuits and bioseparation. In this review, we introduce the fundamental engineering principles used to design RCA nanotechnologies, discuss recently developed RCA-based diagnostics and bioanalytical tools, and summarize the use of RCA to construct multivalent molecular scaffolds and nanostructures for applications in biology, diagnostics and therapeutics.

A Survey of Non-Orthogonal Multiple Access for 5G

Abstract: In the fifth generation (5G) of wireless communication systems, hitherto unprecedented requirements are expected to be satisfied. As one of the promising techniques of addressing these challenges, non-orthogonal multiple access (NOMA) has been actively investigated in recent years. In contrast to the family of conventional orthogonal multiple access (OMA) schemes, the key distinguishing feature of NOMA is to support a higher number of users than the number of orthogonal resource slots with the aid of non-orthogonal resource allocation. This may be realized by the sophisticated inter-user interference cancellation at the cost of an increased receiver complexity. In this paper, we provide a comprehensive survey of the original birth, the most recent development, and the future research directions of NOMA. Specifically, the basic principle of NOMA will be introduced at first, with the comparison between NOMA and OMA especially from the perspective of information theory. Then, the prominent NOMA schemes are discussed by dividing them into two categories, namely, power-domain and code-domain NOMA. Their design principles and key features will be discussed in detail, and a systematic comparison of these NOMA schemes will be summarized in terms of their spectral efficiency, system performance, receiver complexity, etc. Finally, we will highlight a range of challenging open problems that should be solved for NOMA, along with corresponding opportunities and future research trends to address these challenges.

Quantifying global soil carbon losses in response to warming

Abstract: The majority of the Earth's terrestrial carbon is stored in the soil. If anthropogenic warming stimulates the loss of this carbon to the atmosphere, it could drive further planetary warming. Despite evidence that warming enhances carbon fluxes to and from the soil, the net global balance between these responses remains uncertain. Here we present a comprehensive analysis of warming-induced changes in soil carbon stocks by assembling data from 49 field experiments located across North America, Europe and Asia. We find that the effects of warming are contingent on the size of the initial soil carbon stock, with considerable losses occurring in high-latitude areas. By extrapolating this empirical relationship to the global scale, we provide estimates of soil carbon sensitivity to warming that may help to constrain Earth system model projections. Our empirical relationship suggests that global soil carbon stocks in the upper soil horizons will fall by 30 ± 30 petagrams of carbon to 203 ± 161 petagrams of carbon under one degree of warming, depending on the rate at which the effects of warming are realized. Under the conservative assumption that the response of soil carbon to warming occurs within a year, a business-as-usual climate scenario would drive the loss of 55 ± 50 petagrams of carbon from the upper soil horizons by 2050. This value is around 12-17 per cent of the expected anthropogenic emissions over this period. Despite the considerable uncertainty in our estimates, the direction of the global soil carbon response is consistent across all scenarios. This provides strong empirical support for the idea that rising temperatures will stimulate the net loss of soil carbon to the atmosphere, driving a positive land carbon-climate feedback that could accelerate climate change.