List your company

Mouse behavior representation learning on MABe challenge 2022

0.0909Day (MSE)BAMS
0.0908110.0916810.092550.093419Mar 15, 2023

Evaluation results

MethodLinks
BAMSEvaluation protocol=Linear readout, Embedding dimension=128-dim, Pooling=Avg + (Max-Min) concat2023.030.09090.08997288.232.741.8967.2253.4331.430.591.611.5718.15
BAMS long-termEvaluation protocol=Linear readout, Receptive field=1200 frames2023.030.09170.090470.9186.492.11.0661.9949.1229.090.451.321.0813.98
T-BERTEvaluation protocol=Linear readout, Training objective=Masked modeling + Contrastive learning2023.030.09260.092868.8478.631.80.8770.2255.8430.240.511.41.1217.27
T-GPTEvaluation protocol=Linear readout, Training objective=Reconstruction-based2023.030.09270.093865.3964.451.730.6469.0555.7823.80.461.121.0517.86
T-PointNetEvaluation protocol=Linear readout, Training objective=Contrastive learning2023.030.09280.093267.1566.012.564.5770.6855.9621.230.842.792.3215.61
BAMS short-termEvaluation protocol=Linear readout, Receptive field=60 frames2023.030.09290.092966.3461.331.81.1566.5852.625.340.391.090.9816.74
T-PerceiverEvaluation protocol=Linear readout, Training objective=Masked modeling2023.030.09320.093269.6869.811.571.2760.8447.8128.320.411.160.8616.42
TS2VecEvaluation protocol=Linear readout2023.030.09380.094265.657.121.290.6659.5346.1324.740.351.090.7412.37
TVAEEvaluation protocol=Linear readout2023.030.0940.094456.852.981.070.4559.3344.7721.960.270.830.6210.2
PCAEvaluation protocol=Linear readout, Feature set=36 features2023.030.09420.094554.6551.60.860.1449.2737.8712.710.210.60.536.65