List your company
ResearchFormerly WizWand

Multivariate Time Series Classification on LIBRAS (test)

0.97AccuracyMLSTMFCN
0.591440.689720.7880.88628Jun 12, 2020May 28, 2021May 14, 2022Apr 29, 2023Apr 14, 2024Mar 30, 2025Mar 16, 2026
Updated 5mo ago

Evaluation results

MethodLinks
MLSTMFCN2020.060.97
FCN128Backbone=FCN, Filter size=1282020.060.966
RESNETBackbone=ResNet2020.060.966
FCN128-LS2T64Backbone=FCN, Backbone Filter size=128, Rank=642020.060.956
ARKERNEL2020.060.952
FCN64-LS2T64Backbone=FCN, Backbone Filter size=64, Rank=642020.060.946
mvARF2020.060.945
gRSF2020.060.911
SMTS2020.060.909
LPS2020.060.903
MUSE2020.060.894
DTW2020.060.888
LS2T64Rank=642020.060.859
ResRMNIRecurrent Matrix=Identity2025.080.807
RMN2025.080.806
LLM NASm (number of modalities)=1, dm (number of variates)=2, Tm (sequence length)=45, C (number of classes)=15, Train size=180, Test size=180, NAS search budget (cycles)=102026.030.789
Baseline stagedm (number of modalities)=1, dm (number of variates)=2, Tm (sequence length)=45, C (number of classes)=15, Train size=180, Test size=1802026.030.767
ResESNCRecurrent Matrix=Circular2025.080.766
LeakyESN2025.080.76
Previous workm (number of modalities)=1, dm (number of variates)=2, Tm (sequence length)=45, C (number of classes)=15, Train size=180, Test size=1802026.030.756
ResESNIRecurrent Matrix=Identity2025.080.753
ResESNRRecurrent Matrix=Random2025.080.749
ResRMNRRecurrent Matrix=Random2025.080.73
ResRMNCRecurrent Matrix=Circular2025.080.712
Baseline end-to-endm (number of modalities)=1, dm (number of variates)=2, Tm (sequence length)=45, C (number of classes)=15, Train size=180, Test size=1802026.030.606