Zhejiang Gongshang University

About: Zhejiang Gongshang University is a education organization based out in Hangzhou, China. It is known for research contribution in the topics: Adsorption & Supply chain. The organization has 8258 authors who have published 7670 publications receiving 90296 citations. The organization is also known as: Zhèjiāng Gōngshāng Dàxué.

In Situ Method for Real-Time Discriminating Salmon and Rainbow Trout without Sample Preparation Using iKnife and Rapid Evaporative Ionization Mass Spectrometry-Based Lipidomics

Abstract: The domestic rainbow trout producers issued a standard with an aquatic association that classified rainbow trout as salmon, which raised the concern of consumers on the fish parasites infection. Herein, an in situ method was developed using "iKnife" and rapid evaporative ionization mass spectrometry based lipidomics for real-time discrimination of salmon and rainbow trout without sample preparation. A total of 12 fatty acids and 37 phospholipid species was identified and imported into statistical analysis for building an in situ and real-time recognition model. The ions with | p(corr)| > 0.5 and | p| > 0.03 were shown to be responsible for allocating samples, and the ions with high correlation values, such as of m/ z 747.50, 771.49, and 863.55, indicated large weights in identification of the salmon and rainbow trout. The results indicated that this technology could be employed as a front-line test method to ensure the authenticity of salmon products.

Deep Dual Consecutive Network for Human Pose Estimation

Abstract: Multi-frame human pose estimation in complicated situations is challenging. Although state-of-the-art human joints detectors have demonstrated remarkable results for static images, their performances come short when we apply these models to video sequences. Prevalent shortcomings include the failure to handle motion blur, video defocus, or pose occlusions, arising from the inability in capturing the temporal dependency among video frames. On the other hand, directly employing conventional recurrent neural networks incurs empirical difficulties in modeling spatial contexts, especially for dealing with pose occlusions. In this paper, we propose a novel multi-frame human pose estimation framework, leveraging abundant temporal cues between video frames to facilitate keypoint detection. Three modular components are designed in our framework. A Pose Temporal Merger encodes keypoint spatiotemporal context to generate effective searching scopes while a Pose Residual Fusion module computes weighted pose residuals in dual directions. These are then processed via our Pose Correction Network for efficient refining of pose estimations. Our method ranks No.1 in the Multi-frame Person Pose Estimation Challenge on the large-scale benchmark datasets PoseTrack2017 and PoseTrack2018. We have released our code, hoping to inspire future research.