arXiv subjects

Changqian Yu

Publications and source records attributed to Changqian Yu.

39 records · Page 3Linked to original sources

arXiv2018

BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation

Semantic segmentation requires both rich spatial information and sizeable receptive field. However, modern approaches usually compromise spatial resolution to achieve real-time inference speed, which leads to poor performance. In this paper, we address this dilemma with a novel Bilateral Segmentation Network (BiSeNet). We first design a Spatial Path with a small stride to preserve the spatial information and generate high-resolution features. Meanwhile, a Context Path with a fast downsampling strategy is employed to obtain sufficient receptive field. On top of the two paths, we introduce a new Feature Fusion Module to combine features efficiently. The proposed architecture makes a right balance between the speed and segmentation performance on Cityscapes, CamVid, and COCO-Stuff datasets. Specifically, for a 2048x1024 input, we achieve 68.4% Mean IOU on the Cityscapes test dataset with speed of 105 FPS on one NVIDIA Titan XP card, which is significantly faster than the existing methods with comparable performance.

cs.CV↗

arXiv2018

Learning a Discriminative Feature Network for Semantic Segmentation

Most existing methods of semantic segmentation still suffer from two aspects of challenges: intra-class inconsistency and inter-class indistinction. To tackle these two problems, we propose a Discriminative Feature Network (DFN), which contains two sub-networks: Smooth Network and Border Network. Specifically, to handle the intra-class inconsistency problem, we specially design a Smooth Network with Channel Attention Block and global average pooling to select the more discriminative features. Furthermore, we propose a Border Network to make the bilateral features of boundary distinguishable with deep semantic boundary supervision. Based on our proposed DFN, we achieve state-of-the-art performance 86.2% mean IOU on PASCAL VOC 2012 and 80.3% mean IOU on Cityscapes dataset.

cs.CV↗

arXiv2013

Single-Molecule Observation of Long Jumps in Polymer Adsorption

Single-molecule fluorescence imaging of adsorption onto initially-bare surfaces shows that polymer chains need not localize immediately after arrival. In a system optimized to present limited adsorption sites (quartz surface to which polyethylene glycol (PEG) is exposed in aqueous solution at pH = 8.2) we find that some chains diffuse back into bulk solution and re-adsorb at some distance away, sometimes multiple times before either they localize at a stable position or else diffuse away into bulk solution. This mechanism of surface diffusion is considerably more rapid than the classical model in which adsorbed polymers crawl on surfaces while the entire molecule remains adsorbed. The trajectories with jumps follow a truncated Levy distribution of step size with limiting slope -2.5, consistent with a well-defined, rapid surface diffusion coefficient over the times we observe.

cond-mat.soft↗