University of Electronic Science and Technology of China

About: University of Electronic Science and Technology of China is a education organization based out in Chengdu, China. It is known for research contribution in the topics: Antenna (radio) & Dielectric. The organization has 50594 authors who have published 58502 publications receiving 711188 citations. The organization is also known as: UESTC.

Vertical heterostructures based on SnSe2/MoS2 for high performance photodetectors

Abstract: Van der Waals heterostructures from atomically thin 2D materials have opened up new realms in modern semiconductor industry. Recently, 2D layered semiconductors such as MoS2 and SnSe2 have already demonstrated excellent electronic and optoelectronic properties due to their high electron mobility and unique band structures. Such combination of SnSe2 with MoS2 may provide a novel platform for the applications in electronics and optoelectronics. Thus, we constructed SnSe2/MoS2 based van der Waals heterostructures using MoS2 as templates, which may enrich the family of 2D van der Waals heterostructures. We demonstrate that the vdW heterostructures with high symmetry crystallographic directions show efficient interlayer charge transfer due to the strong coupling. This strong coupling is confirmed by theory calculations, low-temperature photoluminescence (PL) spectra, and electrical transport properties. High performance photodetector based on the vdW heterostructure has been demonstrated with a high responsivity of up to 9.1 × 103 A W−1 which is higher by two orders of magnitude than those MoS2-only devices. The improved performance can be attributed to the efficient charge transfer from MoS2 to SnSe2 at the interface.

Anomalous Terahertz Reflection and Scattering by Flexible and Conformal Coding Metamaterials

Abstract: Arbitrary control of terahertz (THz) waves remains a significant challenge although it promises many important applications. Here, a method to tailor the reflection and scattering of THz waves in an anomalous manner by using 1-bit coding metamaterials is presented. Specific coding sequences result in various THz far-field reflection and scattering patterns, ranging from a single beam to two, three, and numerous beams, which depart obviously from the ordinary Snell's law of reflection. By optimizing the coding sequences, a wideband THz thin film metamaterial with extremely low specular reflection, due to the scattering of the incident wave into various directions, is demonstrated. As a result, the reflection from a flat and flexible metamaterial can be nearly uniformly distributed in the half space with small intensity at each specific direction, manifesting a diffuse reflection from a rough surface. Both simulation and experimental results show that a reflectivity less than −10 dB is achieved over a wide frequency range from 0.8 to 1.4 THz, and it is insensitive to the polarization of the incident wave. This work reveals new opportunities arising from coding metamaterials in effective manipulation of THz wave propagation and may offer widespread applications.

Complex synchronous behavior in interneuronal networks with delayed inhibitory and fast electrical synapses.

Abstract: Networks of fast-spiking interneurons are crucial for the generation of neural oscillations in the brain. Here we study the synchronous behavior of interneuronal networks that are coupled by delayed inhibitory and fast electrical synapses. We find that both coupling modes play a crucial role by the synchronization of the network. In addition, delayed inhibitory synapses affect the emerging oscillatory patterns. By increasing the inhibitory synaptic delay, we observe a transition from regular to mixed oscillatory patterns at a critical value. We also examine how the unreliability of inhibitory synapses influences the emergence of synchronization and the oscillatory patterns. We find that low levels of reliability tend to destroy synchronization and, moreover, that interneuronal networks with long inhibitory synaptic delays require a minimal level of reliability for the mixed oscillatory pattern to be maintained.

Attentive Group Recommendation

Abstract: Due to the prevalence of group activities in people's daily life, recommending content to a group of users becomes an important task in many information systems. A fundamental problem in group recommendation is how to aggregate the preferences of group members to infer the decision of a group. Toward this end, we contribute a novel solution, namely AGREE (short for ''Attentive Group REcommEndation''), to address the preference aggregation problem by learning the aggregation strategy from data, which is based on the recent developments of attention network and neural collaborative filtering (NCF). Specifically, we adopt an attention mechanism to adapt the representation of a group, and learn the interaction between groups and items from data under the NCF framework. Moreover, since many group recommender systems also have abundant interactions of individual users on items, we further integrate the modeling of user-item interactions into our method. Through this way, we can reinforce the two tasks of recommending items for both groups and users. By experimenting on two real-world datasets, we demonstrate that our AGREE model not only improves the group recommendation performance but also enhances the recommendation for users, especially for cold-start users that have no historical interactions individually.

Visual Commonsense R-CNN

Abstract: We present a novel unsupervised feature representation learning method, Visual Commonsense Region-based Convolutional Neural Network (VC R-CNN), to serve as an improved visual region encoder for high-level tasks such as captioning and VQA. Given a set of detected object regions in an image (e.g., using Faster R-CNN), like any other unsupervised feature learning methods (e.g., word2vec), the proxy training objective of VC R-CNN is to predict the contextual objects of a region. However, they are fundamentally different: the prediction of VC R-CNN is by using causal intervention: P(Y|do(X)), while others are by using the conventional likelihood: P(Y|X). This is also the core reason why VC R-CNN can learn “sense-making” knowledge like chair can be sat — while not just “common” co-occurrences such as chair is likely to exist if table is observed. We extensively apply VC R-CNN features in prevailing models of three popular tasks: Image Captioning, VQA, and VCR, and observe consistent performance boosts across them, achieving many new state-of-the-arts.