notesum.ai

Published at December 6

KNN-MMD: Cross Domain Wi-Fi Sensing Based on Local Distribution Alignment

cs.CV

cs.AI

eess.SP

Released Date: December 6, 2024

Authors: Zijian Zhao¹, Zhijie Cai, Tingwei Chen, Xiaoyang Li, Hang Li, Guangxu Zhu

Aff.: ¹Shenzhen Research Institute of Big Data, Shenzhen 518115, China, and School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou 510275, China

Arxiv: http://arxiv.org/pdf/2412.04783v1

Method	Scenario	Gesture Recognition	People Identification	Fall Detection	Action Recognition	Average Accuracy
Resnet18[29]	in-domain	80.75%	86.75%	91.88%	70.50%	82.47%
Resnet18[29]	zero-shot	40.84%	70.50%	59.86%	26.00%	49.30%
Siamese[12]	one-shot	70.40%	82.87%	60.62%	38.95%	63.21%
AutoFi (MLP-based)[44]	one-shot	24.62%	24.71%	50.88%	23.59%	30.95%
AutoFi (CNN-based)[44]	one-shot	27.05%	36.14%	48.05%	26.95%	34.55%
Yang et al.[24]	one-shot	67.21%	74.22%	59.75%	48.52%	62.43%
Ding et al.[45]	one-shot	39.14%	70.94%	61.56%	30.37%	50.50%
CrossFi[16]	one-shot	91.72%	93.01%	80.93%	49.62%	78.82%
KNN [30]	one-shot	83.02%	82.67%	49.63%	46.87%	65.55%
KNN-MMD (Ours)	one-shot	93.26%	81.84%	77.62%	75.30%	82.01%
Ablation Study	one-shot	69.87%	73.78%	84.03%	74.06%	75.44%
MMD[17]	zero-shot	47.92%	67.25%	74.32%	45.61%	58.75%
MK-MMD[18]	zero-shot	40.36%	66.47%	72.26%	43.72%	55.70%
DANN[35]	zero-shot	41.41%	67.18%	74.06%	35.99%	54.66%
ADDA[46]	zero-shot	42.71%	65.43%	62.81%	36.08%	51.76%
GFK+KNN[33]	zero-shot	30.79%	51.50%	53.72%	34.17%	42.55%
CrossFi[16]	zero-shot	64.81%	72.79%	74.38%	40.46%	63.11%
Tian et al.[41]	zero-shot	68.13%	55.86%	61.72%	42.10%	56.95%
EEG[42]	zero-shot	59.75%	64.63%	69.53%	42.15%	59.02%