Real valued classification using complex neural networks by Pushkar Shukla.

“…The proposed network will consist of an input layer, which will utilise a circular(sine) function to map the real-valued input onto the complex plane, followed by a hidden layer employing a Gaussian-like sech activation function, followed by the output layer consisting of a single neuron, with encoded outputs corresponding to various class label used to depict the classification of the input data. …”
Get full text

(Dot) Type. by Godfrey, Vanessa Dora.

“…The project’s title follows the popular term “little red dot” that refers to Singapore’s geographical presence on the world map, and was motivated by my training as a visual communicator and my identity as a Singaporean. …”
Get full text

Family social network by Dong, Zeya

“…Photo sharing with story android mobile application provide a platform to share photo with stories and location where the photos are taken with simple features like login process, upload photo on server along with information, face book integration for user comment and likes, Google map integration for location searching where photo has been taken. …”
Get full text

Scene understanding based on heterogeneous data fusion by Zhu, Lingzhi

“…Moreover, after fine tuning process, the highest mAP score it can reach is 0.804. Experiments on testing different variables are implemented, and the results are elaborated as well. …”
Get full text

Implementation of Dijkstra's shortest path algorithm for vehicle routing by Pang, Jacquelyn Jiaqi

“…We will use a traffic simulation software, PTV VISSIM to simulate the traffic conditions along Singapore map. However, to simulate specific conditions, it requires MATLAB (COM programming) to give specific commands to the cars in VISSIM. …”
Get full text

Satellite image fusion for land cover classification by Tsan, Li Ling

“…Through the remote sensing techniques using Synthetic Aperture Radar(SAR), SAR images are pre-processed and later segmented to produce the segmentation maps which gives the land cover classification. To further enhance the land cover classification accuracy, the processed SAR images are fused with the optical images to form a high-resolution composite image. …”
Get full text

Autonomous learning machine on online big data analytics by Lim, Yan Jun

“…One commonly and widely used approach is the so-called MapReduce from Google. Nevertheless, most of existing works are offline in nature, because it expects full access of complete dataset and enable a machine learning algorithm to achieve multiple passes over all data. …”
Get full text

Snake pattern detection algorithm by Yeo, Eugene Han Wei

“…The mobile application includes downloadable map which works offline and alert users when approaching reported snake locations.…”
Get full text

Design of machine learning based face recognition system by Yeo, Alvin Jin Kuang

“…Face recognition is a robust and reliable system which maps out the contour of the person’s face digitally into the computer storing it into the data as a faceprint. …”
Get full text

A temporal cortex cell atlas highlights gene expression dynamics during human brain maturation by Steyn, C, Mishi, R, Fillmore, S, Verhoog, MB, More, J, Rohlwink, UK, Melvill, R, Butler, J, Enslin, JMN, Jacobs, M, Sauka-Spengler, T, Greco, M, Quiñones, S, Dulla, CG, Raimondo, JV, Figaji, A, Hockman, D

“…The resources we present here improve our understanding of the brain during its development and contribute to global efforts to build an inclusive brain cell map.…”

Platelets sequester extracellular DNA, capturing tumour-derived and free foetal DNA by Murphy, L

“…Here I reveal that despite lacking a nucleus, platelets contain a repertoire of DNA fragments that map across the nuclear genome. I identify mechanisms by which platelets sequester DNA and demonstrate foetal DNA in maternal platelets and cancer cell-derived DNA in platelets from patients with pre-malignant lesions and overt cancers. …”

SimSC: a simple framework for semantic correspondence with temperature learning by Li, X, Han, K, Wan, X, Prisacariu, VA

“…We discover that when fine-tuning ImageNet pre-trained backbone on the semantic matching task, L2 normalization of the feature map, a standard procedure in feature matching, produces an overly smooth matching distribution and significantly hinders the fine-tuning process. …”

Bayesian estimation of layers from multiple images by Wexler, Y, Fitzgibbon, A, Zisserman, A

“…Using a Bayesian framework, we model each pixel as a combined sample from the foreground and background and compute a MAP estimate to factor the two. The novelties in this work include the incorporation of three different types of priors for enhancing the results in problematic scenes. …”

Research landscape of energy transition and green finance: A bibliometric analysis by Xu, Jiahui, Liu, Qian, Wider, Walton, Zhang, Shuhan, Muhammad Ashraf, Fauzi, Jiang, Leilei, Udang, Lester Naces, An, Zhida

“…This paper identifies influential publications, maps the research landscape, and forecasts emerging tendencies through co-citation and co-word analyses. …”
Get full text

A modified viscous flow law for natural glacier ice: Scaling from laboratories to ice sheets by Ranganathan, Meghana, Minchew, Brent

“…Here, we present a model of ice deformation that bridges laboratory and glacier scales, unifies existing estimates of the viscous parameters, and provides a framework for estimating the parameters from observations and incorporating flow laws derived from laboratory observations into glacier-flow models. Our results yield a map of the dominant deformation mechanisms in the Antarctic Ice Sheet, showing that, contrary to long-standing assumptions, dislocation creep, characterized by a value of the stress exponent, likely dominates in all fast-flowing areas. …”
Get full text

Deep learning assisted optical wavefront shaping in disordered medium by Luo, Yunqi, Li, Huanhao, Zhang, Ruochong, Lai, Puxiang, Zheng, Yuanjin

“…Simulation results show that the pre-trained neural network is able to map output optical modes to input modes. Compared with previous methods which use iterative optimization, our method realizes a focused speckle pattern with the help of deep learning, which will definitely reduce complexity and time spent in optimization. …”
Get full text

Non-local means improves total-variation constrained photoacoustic image reconstruction by Yalavarthy, Phaneendra K., Kalva, Sandeep Kumar, Pramanik, Manojit, Prakash, Jaya

“…Photoacoustic/Optoacoustic tomography aims to reconstruct maps of the initial pressure rise induced by the absorption of light pulses in tissue. …”
Get full text

Identifying actionable serial correlations in financial markets by Cheong, Siew Ann, Lee, Yann Wei, Li, Ying Ying, Lim, Jia Qing, Tan, Jadie Jiok Duan, Teo, Joan Xin Ping

“…To do so, we downloaded daily closing prices of the Dow Jones Industrial Average component stocks, as well as various assets like stock market indices, United States government bonds, precious metals, commodities, oil and gas, and foreign exchange. We mapped each financial instrument to a letter and their upward movements to words, before testing the frequencies of these words against a null model obtained by reshuffling the empirical time series. …”
Get full text

Direct measurement of a non-Hermitian topological invariant in a hybrid light-matter system by Su, Rui, Estrecho, Eliezer, Biegańska, Dąbrówka, Huang, Yuqing, Wurdack, Matthias, Pieczarka, Maciej, Truscott, Andrew G., Liew, Timothy Chi Hin, Ostrovskaya, Elena A., Xiong, Qihua

“…These are hybrid light-matter quasiparticles formed by photons strongly coupled to electron-hole pairs (excitons) in a halide perovskite semiconductor at room temperature. We experimentally map out both the real (energy) and imaginary (linewidth) parts of the spectrum near the exceptional points and extract the novel topological invariant—fractional spectral winding. …”
Get full text

Towards superior control in automatic face editing with generative adversarial networks by Zhang, Xijue

“…This project examines StyleMapGAN, a novel approach that evolves from StyleGAN by replacing AdaIN with intermediate latent space carrying information on spatial dimensions, hence capable of performing high-quality local editing. …”
Get full text