Learning discriminative and generalizable features with multi-branch for person re-identification

2021 ◽  
pp. 1-15
Author(s):  
Ru Cheng ◽  
Lukun Wang ◽  
Mingrun Wei
Keyword(s):  
Network Architecture ◽  
Global Features ◽  
Spatial Partition ◽  
Function Strategy ◽  
Channel Information ◽  
Previous State ◽  
Multi Level ◽  
Model Training ◽  
The Stability ◽  
Fusion Features

Finer-grained local features play a supplementary role in the description of pedestrian global features, and the combination of them has been an essential solution to improve discriminative performances in person re-identification (PReID) tasks. The existing part-based methods mostly extract representational semantic parts according to human visual habits or some prior knowledge and focus on spatial partition strategies but ignore the significant influence of channel information on PReID task. So, we proposed an end-to-end multi-branch network architecture (MCSN) jointing multi-level global fusion features, channel features and spatial features in this paper to better learn more diverse and discriminative pedestrian features. It is worth noting that the effect of multi-level fusion features on the performance of the model is taken into account when extracting global features. In addition, to enhance the stability of model training and the generalization ability of the model, the BNNeck and the joint loss function strategy are applied to all vector representation branches. Extensive comparative evaluations are conducted on three mainstream image-based evaluation protocols, including Market-1501, DukeMTMC-ReID and MSMT17, to validate the advantages of our proposed model, which outperforms previous state-of-the-art in ReID tasks.

scholarly journals Data-Driven Stability Assessment of Multilayer Long Short-Term Memory Networks

2021 ◽  
Vol 11 (4) ◽  
pp. 1829
Author(s):  
Davide Grande ◽  
Catherine A. Harris ◽  
Giles Thomas ◽  
Enrico Anderlini
Keyword(s):  
Neural Network ◽  
Network Architecture ◽  
Short Term Memory ◽  
Moving Average ◽  
Machine Learning Algorithms ◽  
Physical Models ◽  
Data Driven ◽  
The Stability ◽  
And Control ◽  
Nonlinear Autoregressive

Recurrent Neural Networks (RNNs) are increasingly being used for model identification, forecasting and control. When identifying physical models with unknown mathematical knowledge of the system, Nonlinear AutoRegressive models with eXogenous inputs (NARX) or Nonlinear AutoRegressive Moving-Average models with eXogenous inputs (NARMAX) methods are typically used. In the context of data-driven control, machine learning algorithms are proven to have comparable performances to advanced control techniques, but lack the properties of the traditional stability theory. This paper illustrates a method to prove a posteriori the stability of a generic neural network, showing its application to the state-of-the-art RNN architecture. The presented method relies on identifying the poles associated with the network designed starting from the input/output data. Providing a framework to guarantee the stability of any neural network architecture combined with the generalisability properties and applicability to different fields can significantly broaden their use in dynamic systems modelling and control.

scholarly journals Enabling rapid COVID-19 small molecule drug design through scalable deep learning of generative models

2021 ◽  
pp. 109434202110109
Author(s):  
Sam Ade Jacobs ◽  
Tim Moon ◽  
Kevin McLoughlin ◽  
Derek Jones ◽  
David Hysom ◽  
...  
Keyword(s):  
Small Molecule ◽  
Molecular Design ◽  
Turnaround Time ◽  
Generative Models ◽  
Quality Model ◽  
Training Time ◽  
Small Molecule Drug ◽  
Training Approach ◽  
Previous State ◽  
Multi Level

We improved the quality and reduced the time to produce machine learned models for use in small molecule antiviral design. Our globally asynchronous multi-level parallel training approach strong scales to all of Sierra with up to 97.7% efficiency. We trained a novel, character-based Wasserstein autoencoder that produces a higher quality model trained on 1.613 billion compounds in 23 minutes while the previous state of the art takes a day on 1 million compounds. Reducing training time from a day to minutes shifts the model creation bottleneck from computer job turnaround time to human innovation time. Our implementation achieves 318 PFLOPs for 17.1% of half-precision peak. We will incorporate this model into our molecular design loop enabling the generation of more diverse compounds; searching for novel, candidate antiviral drugs improves and reduces the time to synthesize compounds to be tested in the lab.

scholarly journals A new multi-level algorithm for balanced partition problem on large scale directed graphs

2021 ◽  
Vol 3 (1) ◽  
Author(s):  
Xianyue Li ◽  
Yufei Pang ◽  
Chenxia Zhao ◽  
Yang Liu ◽  
Qingzhen Dong
Keyword(s):  
Large Scale ◽  
Directed Graphs ◽  
Vlsi Design ◽  
Graph Partition ◽  
Partition Problem ◽  
Multi Level ◽  
The Stability ◽  
Recursive Partition ◽  
Balanced Partition ◽  
Partition Method

AbstractGraph partition is a classical combinatorial optimization and graph theory problem, and it has a lot of applications, such as scientific computing, VLSI design and clustering etc. In this paper, we study the partition problem on large scale directed graphs under a new objective function, a new instance of graph partition problem. We firstly propose the modeling of this problem, then design an algorithm based on multi-level strategy and recursive partition method, and finally do a lot of simulation experiments. The experimental results verify the stability of our algorithm and show that our algorithm has the same good performance as METIS. In addition, our algorithm is better than METIS on unbalanced ratio.

ETHNIC AND REGIONAL PARTIES IN THE CEE COUNTRIES: SPECIAL FEATURES AND TRAJECTORIES OF DEVELOPMENT

2020 ◽  
Vol 14 (3) ◽  
pp. 20-34
Author(s):  
Petr Panov ◽  
Keyword(s):  
Ethnic Minority ◽  
Regional Parties ◽  
Ethnic Parties ◽  
Administrative Unit ◽  
Important Player ◽  
Multi Level ◽  
Administrative Units ◽  
The Stability ◽  
Main Factor ◽  
Level Government

In recent decades, in the context of the transformation of national states and the development of multi-level government, there has been an increase in ethnic/regional political parties in Europe. Ethno-regionalism in the CEE countries has a specific basis related to their imperial past, but despite the similarities, each country has special features concerning the strength of parties, their demands and development. The analysis of the most significant ethnic/regional parties in the CEE countries shows that the main factor affecting their strength is the ethnic structure of the population, especially if it is combined with intense ethnic identity, and the ethnic minority has a historical experience of autonomy/statehood. A favorable combination of these factors results in the stability of the electoral strength of ethnic parties, which makes them an important player in the political arena. Concerning the demands of ethnic parties, it has been confirmed that the localization of the respective ethnic minority has a significant effect. If it is in one administrative unit, it stimulates regionalist aspirations; if it dwells in some compactly located administrative units, an ethnic party usually promotes cross-regionalist demands to create a new region. Under conditions of dispersed localization of a minority, an ethnic party does not put forward regionalist claims.

Deep Siamese Metric Learning: A Highly Scalable Approach to Searching Unordered Sets of Trajectories

2022 ◽  
Vol 13 (1) ◽  
pp. 1-23
Author(s):  
Christoffer Löffler ◽  
Luca Reeb ◽  
Daniel Dzibela ◽  
Robert Marzilger ◽  
Nicolas Witt ◽  
...  
Keyword(s):  
Assignment Problem ◽  
Network Architecture ◽  
Metric Learning ◽  
Representation Learning ◽  
Trajectory Data ◽  
Convolutional Network ◽  
Professional Soccer ◽  
Gating Mechanism ◽  
Previous State ◽  
Low Dimensional

This work proposes metric learning for fast similarity-based scene retrieval of unstructured ensembles of trajectory data from large databases. We present a novel representation learning approach using Siamese Metric Learning that approximates a distance preserving low-dimensional representation and that learns to estimate reasonable solutions to the assignment problem. To this end, we employ a Temporal Convolutional Network architecture that we extend with a gating mechanism to enable learning from sparse data, leading to solutions to the assignment problem exhibiting varying degrees of sparsity. Our experimental results on professional soccer tracking data provides insights on learned features and embeddings, as well as on generalization, sensitivity, and network architectural considerations. Our low approximation errors for learned representations and the interactive performance with retrieval times several magnitudes smaller shows that we outperform previous state of the art.

scholarly journals InterNet+: A Light Network for Hand Pose Estimation

Sensors ◽  
2021 ◽  
Vol 21 (20) ◽  
pp. 6747
Author(s):  
Yang Liu ◽  
Jie Jiang ◽  
Jiahao Sun ◽  
Xianghan Wang
Keyword(s):  
Performance Improvement ◽  
Pose Estimation ◽  
Network Architecture ◽  
Activation Function ◽  
Depth Information ◽  
Global Features ◽  
Hand Pose Estimation ◽  
Feature Extractor ◽  
Rgb Images ◽  
Hand Pose

Hand pose estimation from RGB images has always been a difficult task, owing to the incompleteness of the depth information. Moon et al. improved the accuracy of hand pose estimation by using a new network, InterNet, through their unique design. Still, the network still has potential for improvement. Based on the architecture of MobileNet v3 and MoGA, we redesigned a feature extractor that introduced the latest achievements in the field of computer vision, such as the ACON activation function and the new attention mechanism module, etc. Using these modules effectively with our network, architecture can better extract global features from an RGB image of the hand, leading to a greater performance improvement compared to InterNet and other similar networks.

scholarly journals Improving ΔΔG predictions with a multi-task convolutional Siamese Network

2021 ◽  
Author(s):  
Andrew McNutt ◽  
David Koes
Keyword(s):  
Free Energy ◽  
Network Architecture ◽  
Binding Free Energy ◽  
Experimental Testing ◽  
Predictive Performance ◽  
Protein Family ◽  
Siamese Network ◽  
Pearson's R ◽  
Model Training ◽  
Pearson’S R

The lead optimization phase of drug discovery refines an initial hit molecule for desired properties, especially potency. Synthesis and experimental testing of the small perturbations during this refinement can be quite costly and time consuming. Relative binding free energy (RBFE, also referred to as ∆∆G) methods allow the estimation of binding free energy changes after small changes to a ligand scaffold. Here we propose and evaluate a Convolutional Neural Network (CNN) Siamese network for the prediction of RBFE between two bound ligands. We show that our multi-task loss is able to improve on a previous state-of-the-art Siamese network for RBFE prediction via increased regularization of the latent space. The Siamese network architecture is well suited to the prediction of RBFE in comparison to a standard CNN trained on the same data (Pearson’s R of 0.553 and 0.5, respectively). When evaluated on a left-out protein family, our CNN Siamese network shows variability in its RBFE predictive performance depending on the protein family being evaluated (Pearson’s R ranging from-0.44 to 0.97). RBFE prediction performance can be improved during generalization by injecting only a few examples (few-shot learning) from the evaluation dataset during model training.

scholarly journals Mitigation of Torsional Vibrations in Drilling Systems using Adaptive Nonlinear Control System

2020 ◽  
Vol 9 (4) ◽  
pp. 3317-3321
Keyword(s):  
Closed Loop ◽  
Controller Design ◽  
String Model ◽  
Stick Slip ◽  
Rock Interaction ◽  
Performance Specifications ◽  
Level Model ◽  
Multi Level ◽  
The Stability ◽  
Level Training

The reason for this work is to plan a robust yield feedback control way to deal with dispense with torque stick-slip vibrations in boring frameworks. Current industry controllers generally neglect to dispose of stick-slip vibrations, particularly when different torque flex modes assume a job in maniacal assault. In terms of build controller production, a real trainingstring system performs a multi-level model work such as torque mechanics. The proposed controller design is artfully distorted at optimizing the stability with respect to the uncertainty of the nonlinear bit-rock interaction. Based on heroes and intentions. Besides, a closed loop strength examination of the nonlinear preparing string model is displayed. This controller structure system offers a few points of interest contrasted with existing controllers. To begin with, just surface estimations are utilized, barring the requirement for entire estimations underneath it. Second, multi-level training-string dynamics are effectively handled in ways to access state-training controllers. Third, stability is explicitly provided with respect to bit-rock contact uncertainty and closed-loop performance specifications include controller design. The results of the study report confirm that stick-slip vibrations are actually eliminated in realistic drilling scenarios using a controller designed to achieve this state-ofcontrol control.

Multi-Level Radio Network Architecture For Seismic Data Acquisition System

2019 ◽  
Author(s):  
Danis K. Nurgaliev ◽  
Oleg N. Sherstyukov ◽  
Evgeniy Yu. Ryabchenko ◽  
Evgeniy V. Danilov ◽  
Alexey D. Smolyakov ◽  
...  
Keyword(s):  
Data Acquisition ◽  
Seismic Data ◽  
Network Architecture ◽  
Data Acquisition System ◽  
Acquisition System ◽  
Radio Network ◽  
Seismic Data Acquisition ◽  
Multi Level

scholarly journals Research on Face Recognition Method by Autoassociative Memory Based on RNNs

Complexity ◽  
2018 ◽  
Vol 2018 ◽  
pp. 1-12 ◽  
Author(s):  
Qi Han ◽  
Zhengyang Wu ◽  
Shiqin Deng ◽  
Ziqiang Qiao ◽  
Junjian Huang ◽  
...  
Keyword(s):  
Face Recognition ◽  
Design Procedure ◽  
Biological Database ◽  
Model Parameters ◽  
Face Database ◽  
Recognition Method ◽  
The Face ◽  
Model Training ◽  
And Performance ◽  
The Stability

In order to avoid the risk of the biological database being attacked and tampered by hackers, an Autoassociative Memory (AAM) model is proposed in this paper. The model is based on the recurrent neural networks (RNNs) for face recognition, under the condition that the face database is replaced by its model parameters. The stability of the model is proved and analyzed to slack the constraints of AAM model parameters. Besides, a design procedure about solving AAM model parameters is given, and the face recognition method by AAM model is established, which includes image preprocessing, AAM model training, and image recognition. Finally, simulation results on two experiments show the feasibility and performance of the proposed face recognition method.

Sign in / Sign up

Export Citation Format

Share Document