Abstract:Virtual reality (VR) presents immersive opportunities across many applications, yet the inherent risk of developing cybersickness during interaction can severely reduce enjoyment and platform adoption. Cybersickness is marked by symptoms such as dizziness and nausea, which previous work primarily assessed via subjective post-immersion questionnaires and motion-restricted controlled setups. In this paper, we investigate the \emph{dynamic nature} of cybersickness while users experience and freely interact in VR. We propose a novel method to \emph{continuously} identify and quantitatively gauge cybersickness levels from users' \emph{passively monitored} electroencephalography (EEG) and head motion signals. Our method estimates multitaper spectrums from EEG, integrating specialized EEG processing techniques to counter motion artifacts, and, thus, tracks cybersickness levels in real-time. Unlike previous approaches, our method requires no user-specific calibration or personalization for detecting cybersickness. Our work addresses the considerable challenge of reproducibility and subjectivity in cybersickness research.
Abstract:High-quality, error-free datasets are a key ingredient in building reliable, accurate, and unbiased machine learning (ML) models. However, real world datasets often suffer from errors due to sensor malfunctions, data entry mistakes, or improper data integration across multiple sources that can severely degrade model performance. Detecting and correcting these issues typically require tailor-made solutions and demand extensive domain expertise. Consequently, automation is challenging, rendering the process labor-intensive and tedious. In this study, we investigate whether Large Language Models (LLMs) can help alleviate the burden of manual data cleaning. We set up an experiment in which an LLM, paired with Python, is tasked with cleaning the training dataset to improve the performance of a learning algorithm without having the ability to modify the training pipeline or perform any feature engineering. We run this experiment on multiple Kaggle datasets that have been intentionally corrupted with errors. Our results show that LLMs can identify and correct erroneous entries, such as illogical values or outlier, by leveraging contextual information from other features within the same row, as well as feedback from previous iterations. However, they struggle to detect more complex errors that require understanding data distribution across multiple rows, such as trends and biases.
Abstract:Egocentric vision systems aim to understand the spatial surroundings and the wearer's behavior inside it, including motions, activities, and interaction with objects. Since a person's attention and situational responses are influenced by their physiological state, egocentric systems must also detect this state for better context awareness. In this paper, we propose egoPPG, a novel task for egocentric vision systems to extract a person's heart rate (HR) as a key indicator of the wearer's physiological state from the system's built-in sensors (e.g., eye tracking videos). We then propose EgoPulseFormer, a method that solely takes eye-tracking video as input to estimate a person's photoplethysmogram (PPG) from areas around the eyes to track HR values-without requiring additional or dedicated hardware. We demonstrate the downstream benefit of EgoPulseFormer on EgoExo4D, where we find that augmenting existing models with tracked HR values improves proficiency estimation by 14%. To train and validate EgoPulseFormer, we collected a dataset of 13+ hours of eye-tracking videos from Project Aria and contact-based blood volume pulse signals as well as an electrocardiogram (ECG) for ground-truth HR values. 25 participants performed diverse everyday activities such as office work, cooking, dancing, and exercising, which induced significant natural motion and HR variation (44-164 bpm). Our model robustly estimates HR (MAE=8.82 bpm) and captures patterns (r=0.81). Our results show how egocentric systems may unify environmental and physiological tracking to better understand user actions and internal states.
Abstract:Deep learning models lack shift invariance, making them sensitive to input shifts that cause changes in output. While recent techniques seek to address this for images, our findings show that these approaches fail to provide shift-invariance in time series, where the data generation mechanism is more challenging due to the interaction of low and high frequencies. Worse, they also decrease performance across several tasks. In this paper, we propose a novel differentiable bijective function that maps samples from their high-dimensional data manifold to another manifold of the same dimension, without any dimensional reduction. Our approach guarantees that samples -- when subjected to random shifts -- are mapped to a unique point in the manifold while preserving all task-relevant information without loss. We theoretically and empirically demonstrate that the proposed transformation guarantees shift-invariance in deep learning models without imposing any limits to the shift. Our experiments on six time series tasks with state-of-the-art methods show that our approach consistently improves the performance while enabling models to achieve complete shift-invariance without modifying or imposing restrictions on the model's topology. The source code is available on \href{https://github.com/eth-siplab/Shifting-the-Paradigm}{GitHub}.
Abstract:Research on egocentric tasks in computer vision has mostly focused on head-mounted cameras, such as fisheye cameras or embedded cameras inside immersive headsets. We argue that the increasing miniaturization of optical sensors will lead to the prolific integration of cameras into many more body-worn devices at various locations. This will bring fresh perspectives to established tasks in computer vision and benefit key areas such as human motion tracking, body pose estimation, or action recognition -- particularly for the lower body, which is typically occluded. In this paper, we introduce EgoSim, a novel simulator of body-worn cameras that generates realistic egocentric renderings from multiple perspectives across a wearer's body. A key feature of EgoSim is its use of real motion capture data to render motion artifacts, which are especially noticeable with arm- or leg-worn cameras. In addition, we introduce MultiEgoView, a dataset of egocentric footage from six body-worn cameras and ground-truth full-body 3D poses during several activities: 119 hours of data are derived from AMASS motion sequences in four high-fidelity virtual environments, which we augment with 5 hours of real-world motion data from 13 participants using six GoPro cameras and 3D body pose references from an Xsens motion capture suit. We demonstrate EgoSim's effectiveness by training an end-to-end video-only 3D pose estimation network. Analyzing its domain gap, we show that our dataset and simulator substantially aid training for inference on real-world data. EgoSim code & MultiEgoView dataset: https://siplab.org/projects/EgoSim
Abstract:A requirement of cross-modal signal processing is accurate signal alignment. Though simple on a single device, accurate signal synchronization becomes challenging as soon as multiple devices are involved, such as during activity monitoring, health tracking, or motion capture - particularly outside controlled scenarios where data collection must be standalone, low-power, and support long runtimes. In this paper, we present BMAR, a novel synchronization method that operates purely based on recorded signals and is thus suitable for offline processing. BMAR needs no wireless communication between devices during runtime and does not require any specific user input, action, or behavior. BMAR operates on the data from devices worn by the same person that record barometric pressure and acceleration - inexpensive, low-power, and thus commonly included sensors in today's wearable devices. In its first stage, BMAR verifies that two recordings were acquired simultaneously and pre-aligns all data traces. In a second stage, BMAR refines the alignment using acceleration measurements while accounting for clock skew between devices. In our evaluation, three to five body-worn devices recorded signals from the wearer for up to ten hours during a series of activities. BMAR synchronized all signal recordings with a median error of 33.4 ms and reliably rejected non-overlapping signal traces. The worst-case activity was sleeping, where BMAR's second stage could not exploit motion for refinement and, thus, aligned traces with a median error of 3.06 s.
Abstract:Emotions are known to mediate the relationship between users' content consumption and their online engagement, with heightened emotional intensity leading to increased engagement. Building on this insight, we propose three regressor-guided image editing approaches aimed at diminishing the emotional impact of images. These include (i) a parameter optimization approach based on global image transformations known to influence emotions, (ii) an optimization approach targeting the style latent space of a generative adversarial network, and (iii) a diffusion-based approach employing classifier guidance and classifier-free guidance. Our findings demonstrate that approaches can effectively alter the emotional properties of images while maintaining high visual quality. Optimization-based methods primarily adjust low-level properties like color hues and brightness, whereas the diffusion-based approach introduces semantic changes, such as altering appearance or facial expressions. Notably, results from a behavioral study reveal that only the diffusion-based approach successfully elicits changes in viewers' emotional responses while preserving high perceived image quality. In future work, we will investigate the impact of these image adaptations on internet user behavior.
Abstract:Reflective photoplethysmography (PPG) has become the default sensing technique in wearable devices to monitor cardiac activity via a person's heart rate (HR). However, PPG-based HR estimates can be substantially impacted by factors such as the wearer's activities, sensor placement and resulting motion artifacts, as well as environmental characteristics such as temperature and ambient light. These and other factors can significantly impact and decrease HR prediction reliability. In this paper, we show that state-of-the-art HR estimation methods struggle when processing \emph{representative} data from everyday activities in outdoor environments, likely because they rely on existing datasets that captured controlled conditions. We introduce a novel multimodal dataset and benchmark results for continuous PPG recordings during outdoor activities from 16 participants over 13.5 hours, captured from four wearable sensors, each worn at a different location on the body, totaling 216\,hours. Our recordings include accelerometer, temperature, and altitude data, as well as a synchronized Lead I-based electrocardiogram for ground-truth HR references. Participants completed a round trip from Zurich to Jungfraujoch, a tall mountain in Switzerland over the course of one day. The trip included outdoor and indoor activities such as walking, hiking, stair climbing, eating, drinking, and resting at various temperatures and altitudes (up to 3,571\,m above sea level) as well as using cars, trains, cable cars, and lifts for transport -- all of which impacted participants' physiological dynamics. We also present a novel method that estimates HR values more robustly in such real-world scenarios than existing baselines.
Abstract:Smartwatches have become popular for monitoring physiological parameters outside clinical settings. Using reflective photoplethysmography (PPG) sensors, such watches can non-invasively estimate heart rate (HR) in everyday environments and throughout a patient's day. However, achieving consistently high accuracy remains challenging, particularly during moments of increased motion or due to varying device placement. In this paper, we introduce a novel sensor fusion method for estimating HR that flexibly combines samples from multiple PPG sensors placed across the patient's body, including wrist, ankle, head, and sternum (chest). Our method first estimates signal quality across all inputs to dynamically integrate them into a joint and robust PPG signal for HR estimation. We evaluate our method on a novel dataset of PPG and ECG recordings from 14 participants who engaged in real-world activities outside the laboratory over the course of a whole day. Our method achieves a mean HR error of 2.4\,bpm, which is 46\% lower than the mean error of the best-performing single device (4.4\,bpm, head).
Abstract:Multi-channel photoplethysmography (PPG) sensors have found widespread adoption in wearable devices for monitoring cardiac health. Channels thereby serve different functions -- whereas green is commonly used for metrics such as heart rate and heart rate variability, red and infrared are commonly used for pulse oximetry. In this paper, we introduce a novel method that simultaneously fuses multi-channel PPG signals into a single recovered PPG signal that can be input to further processing. Via signal fusion, our learning-based method compensates for the artifacts that affect wavelengths to different extents, such as motion and ambient light changes. We evaluate our method on a novel dataset of multi-channel PPG recordings and electrocardiogram recordings for reference from 10 participants over the course of 13 hours during real-world activities outside the laboratory. Using the fusion PPG signal our method recovered, participants' heart rates can be calculated with a mean error of 4.5\,bpm (23\% lower than from green PPG signals at 5.9\,bpm).