Picture for Weixin Yang

Weixin Yang

MCGAN: Enhancing GAN Training with Regression-Based Generator Loss

Add code
May 27, 2024
Viaarxiv icon

GCN-DevLSTM: Path Development for Skeleton-Based Action Recognition

Add code
Mar 22, 2024
Figure 1 for GCN-DevLSTM: Path Development for Skeleton-Based Action Recognition
Figure 2 for GCN-DevLSTM: Path Development for Skeleton-Based Action Recognition
Figure 3 for GCN-DevLSTM: Path Development for Skeleton-Based Action Recognition
Figure 4 for GCN-DevLSTM: Path Development for Skeleton-Based Action Recognition
Viaarxiv icon

Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition

Add code
Nov 01, 2021
Figure 1 for Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition
Figure 2 for Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition
Figure 3 for Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition
Figure 4 for Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition
Viaarxiv icon

Computing the full signature kernel as the solution of a Goursat problem

Add code
Jun 26, 2020
Figure 1 for Computing the full signature kernel as the solution of a Goursat problem
Figure 2 for Computing the full signature kernel as the solution of a Goursat problem
Viaarxiv icon

Learning stochastic differential equations using RNN with log signature features

Add code
Sep 22, 2019
Figure 1 for Learning stochastic differential equations using RNN with log signature features
Figure 2 for Learning stochastic differential equations using RNN with log signature features
Figure 3 for Learning stochastic differential equations using RNN with log signature features
Figure 4 for Learning stochastic differential equations using RNN with log signature features
Viaarxiv icon

Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module

Add code
Dec 07, 2018
Figure 1 for Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module
Figure 2 for Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module
Figure 3 for Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module
Figure 4 for Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module
Viaarxiv icon

Leveraging the Path Signature for Skeleton-based Human Action Recognition

Add code
Jul 13, 2017
Figure 1 for Leveraging the Path Signature for Skeleton-based Human Action Recognition
Figure 2 for Leveraging the Path Signature for Skeleton-based Human Action Recognition
Figure 3 for Leveraging the Path Signature for Skeleton-based Human Action Recognition
Figure 4 for Leveraging the Path Signature for Skeleton-based Human Action Recognition
Viaarxiv icon

Online Signature Verification using Recurrent Neural Network and Length-normalized Path Signature

Add code
May 19, 2017
Figure 1 for Online Signature Verification using Recurrent Neural Network and Length-normalized Path Signature
Figure 2 for Online Signature Verification using Recurrent Neural Network and Length-normalized Path Signature
Figure 3 for Online Signature Verification using Recurrent Neural Network and Length-normalized Path Signature
Figure 4 for Online Signature Verification using Recurrent Neural Network and Length-normalized Path Signature
Viaarxiv icon

Building Fast and Compact Convolutional Neural Networks for Offline Handwritten Chinese Character Recognition

Add code
Feb 26, 2017
Figure 1 for Building Fast and Compact Convolutional Neural Networks for Offline Handwritten Chinese Character Recognition
Figure 2 for Building Fast and Compact Convolutional Neural Networks for Offline Handwritten Chinese Character Recognition
Figure 3 for Building Fast and Compact Convolutional Neural Networks for Offline Handwritten Chinese Character Recognition
Figure 4 for Building Fast and Compact Convolutional Neural Networks for Offline Handwritten Chinese Character Recognition
Viaarxiv icon

Toward high-performance online HCCR: a CNN approach with DropDistortion, path signature and spatial stochastic max-pooling

Add code
Feb 24, 2017
Figure 1 for Toward high-performance online HCCR: a CNN approach with DropDistortion, path signature and spatial stochastic max-pooling
Figure 2 for Toward high-performance online HCCR: a CNN approach with DropDistortion, path signature and spatial stochastic max-pooling
Figure 3 for Toward high-performance online HCCR: a CNN approach with DropDistortion, path signature and spatial stochastic max-pooling
Figure 4 for Toward high-performance online HCCR: a CNN approach with DropDistortion, path signature and spatial stochastic max-pooling
Viaarxiv icon