Neural Architecture Transfer

Neural architecture search (NAS) has emerged as a promising avenue for automatically designing task-specific neural networks. Existing NAS approaches require one complete search for each deployment…

Explicit Filterbank Learning for Neural Image Style Transfer and Image Proces…

Multiobjective Evolutionary Design of Deep Convolutional Neural Networks for …

Convolutional neural networks (CNNs) are the backbones of deep learning paradigms for numerous vision tasks. Early advancements in CNN architectures are primarily driven by human expertise and by e…

Population and environment : methodes of analysis

Frontiers of population forecasting

Side-Polished Single-Mode-Multimode-Single-Mode Fiber Structure for the Vecto…

A vector magnetometer, composed of a side-polished single-mode-multimode-single-mode fiber structure coated with ferrofluids, is proposed and investigated. The side-polishing of the multimode fiber…

Pengantar sosiologi : buku panduan mahasiswa

Birefringence Measurement by Expandable Polarization Interference Method

Unrepeatered Transmission Over 670.64 km of 50G BPSK, 653.35 km of 100G PS-QP…

Modeling Gain Mechanisms in Amorphous Silicon Due to Efficient Carrier Multip…

Amorphous materials have low mobility due to their nature of disorder. Surprisingly, some disordered materials showed photocurrent amplification not by conventional photoconductive gain. Recently, …