In recent years, heterogeneous integration (HI) has become a game-changing technology for the construction of complex photonic integrated circuits. Comparing to monolithic integration (MI) techniqu…
This paper introduces versatile filters to construct efficient convolutional neural networks that are widely used in various visual recognition tasks. Considering the demands of efficient deep lear…
Feature reassembly, i.e. feature downsampling and upsampling, is a key operation in a number of modern convolutional network architectures, e.g., residual networks and feature pyramids. Its design …
We focus on a fundamental task of detecting meaningful line structures, a.k.a. , semantic line, in natural scenes. Many previous methods regard this problem as a special case of object detection an…
We propose and experimentally demonstrate a slit beam shaping method for femtosecond laser Point-by-Point (PbP) inscription of highly localized Fiber Bragg Gratings (FBGs). The influence of slit wi…
In neural networks, developing regularization algorithms to settle overfitting is one of the major study areas. We propose a new approach for the regularization of neural networks by the local Rade…
The generator in generative adversarial networks (GANs) is driven by a discriminator to produce high-quality images through an adversarial game. At the same time, the difficulty of reaching a stabl…
We propose a temperature sensor, which consists of a composite optical waveguide (COW) sandwiched between two single mode fibers (SMFs). The COW is formed by side polishing a capillary optical fibe…
An optical vector analyzer assisted by stimulated Brillouin scattering (SBS) processing is proposed and experimentally demonstrated. Benefitting from the selective amplification and attenuation ach…
The key idea of this work is to treat the symbol- pattern-dependent inter-symbol-interference (ISI) that is avoided in conventional single-symbol modulation systems as useful symbol correlation, in…