Related Experiment Video
Updated: Jul 15, 2026
08:25
Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
Published on: May 7, 2019
9.0K
Use of Domain Labels during Pre-Training for Domain-Independent WiFi-CSI Gesture Recognition.
Bram van Berlo1, Richard Verhoeven1, Nirvana Meratnia1
1Department of Mathematics and Computer Science, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands.
Sensors (Basel, Switzerland)
|November 25, 2023
Summary
This study introduces adversarial domain classification to improve WiFi-CSI gesture recognition by reducing domain shift during unsupervised pre-training. Multi-label domain classification effectively minimizes domain variations in position and orientation.
Area of Science:
- Computer Science
- Machine Learning
- Signal Processing
Background:
- Unsupervised representation learning is crucial for WiFi-CSI gesture recognition to minimize reliance on labeled data.
- Domain factors in WiFi-CSI data negatively impact pre-training performance, hindering downstream task accuracy.
- Existing methods struggle to mitigate the effects of domain shifts during the unsupervised pre-training phase.
Purpose of the Study:
- To propose and evaluate the integration of adversarial domain classification within the pre-training phase of WiFi-CSI gesture recognition.
- To investigate the effectiveness of multi-class and label versions of domain classification in reducing domain impact.
- To enhance automatic domain discovery during unsupervised representation learning.
Main Methods:
- Implemented an adversarial domain classifier integrated into the unsupervised representation learning pre-training phase.
- Conducted extensive cross-validation experiments using Widar3 and SignFi datasets with random and leave-out domain factors.
- Evaluated both single- and multi-label domain classification strategies and a domain-aware unsupervised baseline.
Main Results:
- Multi-label domain classification demonstrated a reduction in domain shift for position (1.2% mean improvement) and orientation (0.4% mean improvement) on the Widar3 dataset.
- Domain shift reduction was negatively impacted when negative view combinations spanned multiple domains during pre-training.
- View contrastive loss, when repelling diverse domain views, exacerbated domain shift in the feature space.
Conclusions:
- Adversarial domain classification, particularly multi-label, is an effective strategy to mitigate domain shift in WiFi-CSI gesture recognition pre-training.
- Careful consideration of negative view sampling is necessary to prevent increased domain shift during contrastive learning.
- The proposed method advances unsupervised learning for robust human-computer interaction using WiFi-CSI data.
Related Concept Videos
03:31Labeling DNA Probes
DNA probes are fragments of DNA labeled with a reporter tag to enable their detection or purification. The resulting labeled DNA probes can then hybridize to target nucleic acid sequences through complementary base-pairing, and may be used to recover or identify these regions.
Radioisotopes, fluorophores, or small molecule binding partners like biotin or digoxigenin, are the most widely used reporter tags for labeling DNA probes. These labels can be attached to the probe DNA molecule via...
Radioisotopes, fluorophores, or small molecule binding partners like biotin or digoxigenin, are the most widely used reporter tags for labeling DNA probes. These labels can be attached to the probe DNA molecule via...
01:03IR Frequency Region: Fingerprint Region
IR spectra are divided into two main regions: the diagnostic region and the fingerprint region. The diagnostic region of the spectrum lies above 1500 cm−1. The absorptions resulting from single-bond vibrations of the N–H, C–H, and O–H stretch at higher wavenumbers and appear on the left side of the spectrum. The stretching absorptions of the C≡C and C≡N occur between 2100–2300 cm−1. In contrast, those arising from stretching absorptions of the C=O, C=N, and C=C occur between 1600–1850 cm−1.
The...
The...
01:22Force Classification
Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
01:20Labeling Emotion
Emotional labeling is a cognitive process that involves identifying and naming one's emotions, such as anger, fear, happiness, or sadness. It allows individuals to recognize and express their internal emotional states, a critical aspect of emotional regulation and communication. Labeling emotions requires more than mere recognition; it also involves drawing upon memory and contextual cues to understand the current situation and apply a corresponding emotional label. For instance, feeling...