arXiv subjects

Yutong Feng

Publications and source records attributed to Yutong Feng.

42 records · Page 3Linked to original sources

arXiv2021

View-Guided Point Cloud Completion

This paper presents a view-guided solution for the task of point cloud completion. Unlike most existing methods directly inferring the missing points using shape priors, we address this task by introducing ViPC (view-guided point cloud completion) that takes the missing crucial global structure information from an extra single-view image. By leveraging a framework that sequentially performs effective cross-modality and cross-level fusions, our method achieves significantly superior results over typical existing solutions on a new large-scale dataset we collect for the view-guided point cloud completion task.

cs.CV↗

arXiv2020

High power, electronically-controlled, source of user-defined vortex and vector light beams based on a few-mode fibre amplifier

Orbital angular momentum (OAM) based structured light beams provide an additional degree of freedom for practical applications ranging from optical communication to laser-based material processing. Many techniques exist for generating such beams within laser sources and these primarily rely upon the use of specially designed optical components that limit laser power scaling and ready tunability of the topological charge and polarization of the output OAM beams. Here we show that some of these limitations can be overcome by employing a computer controlled reflective phase-only spatial light modulator (SLM) to adaptively tailor the input (and subsequent output) beam wavefront and polarization in a few-mode fibre amplifier. In this way modal-coupling induced beam distortion within the fibre amplifier can be mitigated and we are able to generate at will any desired supported spatial mode guided in the fibre, including conventional LP modes, scalar OAM modes and cylindrical vector modes, at average powers >10 W and with a peak power of >11 kW. Our results pave the way to the realization of practical high-power structured laser sources with tunable chirality and polarization.

physics.optics↗

arXiv2020

Efficient extraction of high pulse energy from partly quenched highly Er3+-doped fiber amplifiers

We demonstrate efficient pulse-energy extraction from a partly-quenched erbium-doped aluminosilicate fiber amplifier. This has a high erbium-concentration, which allows for short devices with reduced nonlinear distortions, but which also results in partial quenching and thus significant unsaturable absorption, even though the fiber is still able to amplify. Although the quenching degrades the average-power efficiency, the pulse energy remains high, and our results point to an increasingly promising outcome for short pulses. Furthermore, unlike unquenched fibers, the conversion efficiency improves at low repetition rates, which we attribute to smaller relative energy loss to quenched ions at higher pulse energy. A short (2.6 m) cladding-pumped partly-quenched Er-doped-fiber with 95-dB/m 1530-nm peak absorption and saturation energy estimated to 85 μJ, reached 0.8 mJ of output energy when seeded by 0.2-μs, 23-μJ pulses. Thus, according to our results, pulses can be amplified to high energy in short highly-Er-doped fibers designed to reduce nonlinear distortions, at the expense of average-power efficiency

physics.optics↗

arXiv2020

Drug dissemination strategy with an SEIR-based SUC model

According to the features of drug addiction, this paper constructs an SEIR-based SUC model to describe and predict the spread of drug addiction. Predictions are that the number of drug addictions will continue to fluctuate with reduced amplitude and eventually stabilize. To seek the fountainhead of heroin, we identified the most likely origins of drugs in Philadelphia, PA, Cuyahoga and Hamilton, OH, Jefferson, KY, Kanawha, WV, and Bedford, VA. Based on the facts, advised concentration includes the spread of Oxycodone, Hydrocodone, Heroin, and Buprenorphine. In other words, drug transmission in the two states of Ohio and Pennsylvania require awareness. According to the propagation curve predicted by our model, the transfer of KY state is still in its early stage, while that of VA, WV is in the middle point, and OH, PA in its latter ones. As a result of this, the number of drug addictions in KY, OH, and VA is projected to increase in three years. For methodology, with the Principal component analysis technique, 22 variables in socio-economic data related to the continuous use of Opioid drugs was filtered, where the 'Relationship' Part deserves a highlight. Based on them, by using the K-means algorithm, 464 counties were categorized into three baskets. To combat the opioid crisis, a specific action will discuss in the sensitivity analysis section. After modeling and analytics, innovation is required to control addicts and advocate anti-drug news campaigns. This part also verified the effectiveness of model when $d_1<0.2; r_1,r_2,r_3<0.3; 15<β_1,β_2,β_3<25$. In other words, if such boundary exceeded, the number of drug addictions may rocket and peak in a short period.

q-bio.PE↗

arXiv2019

Design of High-Frequency Trading Algorithm Based on Machine Learning

Based on iterative optimization and activation function in deep learning, we proposed a new analytical framework of high-frequency trading information, that reduced structural loss in the assembly of Volume-synchronized probability of Informed Trading ($VPIN$), Generalized Autoregressive Conditional Heteroscedasticity (GARCH) and Support Vector Machine (SVM) to make full use of the order book information. Amongst the return acquisition procedure in market-making transactions, uncovering the relationship between discrete dimensional data from the projection of high-dimensional time-series would significantly improve the model effect. $VPIN$ would prejudge market liquidity, and this effectiveness backtested with CSI300 futures return.

q-fin.TR↗

arXiv2018

MeshNet: Mesh Neural Network for 3D Shape Representation

Mesh is an important and powerful type of data for 3D shapes and widely studied in the field of computer vision and computer graphics. Regarding the task of 3D shape representation, there have been extensive research efforts concentrating on how to represent 3D shapes well using volumetric grid, multi-view and point cloud. However, there is little effort on using mesh data in recent years, due to the complexity and irregularity of mesh data. In this paper, we propose a mesh neural network, named MeshNet, to learn 3D shape representation from mesh data. In this method, face-unit and feature splitting are introduced, and a general architecture with available and effective blocks are proposed. In this way, MeshNet is able to solve the complexity and irregularity problem of mesh and conduct 3D shape representation well. We have applied the proposed MeshNet method in the applications of 3D shape classification and retrieval. Experimental results and comparisons with the state-of-the-art methods demonstrate that the proposed MeshNet can achieve satisfying 3D shape classification and retrieval performance, which indicates the effectiveness of the proposed method on 3D shape representation.

cs.CV↗