Author name cluster

Xiaobai Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

AAAI Conference 2026 Conference Paper

Radar-APLANC: Unsupervised Radar-based Heartbeat Sensing via Augmented Pseudo-Label and Noise Contrast

Ying Wang
Zhaodong Sun
Xu Cheng
Zuxian He
Xiaobai Li

Frequency Modulated Continuous Wave (FMCW) radars can measure subtle chest wall oscillations to enable non-contact heartbeat sensing. However, traditional radar-based heartbeat sensing methods face performance degradation due to noise. Learning-based radar methods achieve better noise robustness but require costly labeled signals for supervised training. To overcome these limitations, we propose the first unsupervised framework for radar-based heartbeat sensing via Augmented Pseudo-Label and Noise Contrast (Radar-APLANC). We propose to use both the heartbeat range and noise range within the radar range matrix to construct the positive and negative samples, respectively, for improved noise robustness. Our Noise-Contrastive Triplet (NCT) loss only utilizes positive samples, negative samples, and pseudo-label signals generated by the traditional radar method, thereby avoiding dependence on expensive ground-truth physiological signals. We further design a pseudo-label augmentation approach featuring adaptive noise-aware label selection to improve pseudo-label signal quality. Extensive experiments on the Equipleth dataset and our collected radar dataset demonstrate that our unsupervised method achieves performance comparable to state-of-the-art supervised methods.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

Active Multimodal Distillation for Few-shot Action Recognition

Weijia Feng
Yichen Zhu
Ruojia Zhang
Chenyang Wang
Fei Ma
Xiaobao Wang
Xiaobai Li

Owing to its rapid progress and broad application prospects, few-shot action recognition has attracted considerable interest. However, current methods are predominantly based on limited single-modal data, which does not fully exploit the potential of multimodal information. This paper presents a novel framework that actively identifies reliable modalities for each sample using task-specific contextual cues, thus significantly improving recognition performance. Our framework integrates an Active Sample Inference (ASI) module, which utilizes active inference to predict reliable modalities based on posterior distributions and subsequently organizes them accordingly. Unlike reinforcement learning, active inference replaces rewards with evidence-based preferences, making more stable predictions. Additionally, we introduce an active mutual distillation module that enhances the representation learning of less reliable modalities by transferring knowledge from more reliable ones. Adaptive multimodal inference is employed during the meta-test to assign higher weights to reliable modalities. Extensive experiments across multiple benchmarks demonstrate that our method significantly outperforms existing approaches.

PDF Details DOI

AAAI Conference 2025 Conference Paper

PhysDiff: Physiology-based Dynamicity Disentangled Diffusion Model for Remote Physiological Measurement

Wei Qian
Gaoji Su
Dan Guo
Jinxing Zhou
Xiaobai Li
Bin Hu
Shengeng Tang
Meng Wang

Recent works on remote PhotoPlethysmoGraphy (rPPG) estimation typically use techniques like CNNs and Transformers to encode implicit features from facial videos for prediction. These methods learn to directly map facial videos to the static values of rPPG signals, overlooking the inherent dynamic characteristics of rPPG sequence. Moreover, the rPPG signal is extremely weak and highly susceptible to interference from various sources of noise, including illumination conditions, head movements, and variations in skin tone. To address these limitations, we propose a Physiology-based dynamicity disentangled diffusion (PhysDiff) model particularly designed for robust rPPG estimation. PhysDiff leverages the diffusion model to learn the distribution of quasi-periodic rPPG signal and uses a dynamicity disentanglement strategy to capture two dynamic characteristics in temporal rPPG signal, i.e., trend and amplitude. This disentanglement is motivated by the underlying dynamic physiological processes of vasodilation and vasoconstriction, ensuring a more precise representation of the rPPG signal. The disentangled components are then used as pivotal conditions in the proposed spatial-temporal hybrid denoiser for rPPG reconstruction. Besides, we introduce a periodicity-based multi-hypothesis selection strategy in model inference, which compares the natural periodicity of multiple generated rPPG hypotheses and selects the most favorable one as the final prediction. Extensive experiments on four datasets demonstrate that our PhysDiff significantly outperforms prior methods on both intra-dataset and cross-dataset testing.

PDF Details DOI

ECAI Conference 2024 Conference Paper

Exemplar-Free Incremental Deepfake Detection

Wuti Xiong
Guoying Zhao 0001
Xiaobai Li

Incremental Deepfake Detection (IDD) aims to continuously update models with new domain data, adapting to evolving forgery techniques. Existing works require extra buffers to store old exemplars for maintaining previously learned knowledge. However, it is infeasible when previous data is unavailable due to storage and privacy issues. This paper focuses on a more challenging but practical exemplar-free IDD problem that requests zero old exemplars when updating the model. To address this problem, we design a domain-adaptive module that uses independent adapters to learn domain-specific knowledge for each domain, avoiding using old exemplars. Besides, we introduce an uncertainty optimization strategy to optimize the adapters more efficiently. With excellent scalability, our method can be easily deployed to various models. To simulate the practical scenarios, we designed two new protocols based on diverse deepfake datasets. Extensive experimental results demonstrate that our method outperforms the state-of-the-art methods by a large margin. The code is available at https: //github. com/woody-panda/EF-IDD.

Details

JBHI Journal 2024 Journal Article

HRUNet: Assessing Uncertainty in Heart Rates Measured From Facial Videos

Xuenan Liu
Xuezhi Yang
Xiaobai Li

Video-based Photoplethysmography (VPPG) offers the capability to measure heart rate (HR) from facial videos. However, the reliability of the HR values extracted through this method remains uncertain, especially when videos are affected by various disturbances. Confronted by this challenge, we introduce an innovative framework for VPPG-based HR measurements, with a focus on capturing diverse sources of uncertainty in the predicted HR values. In this context, a neural network named HRUNet is structured for HR extraction from input facial videos. Departing from the conventional training approach of learning specific weight (and bias) values, we leverage the Bayesian posterior estimation to derive weight distributions within HRUNet. These distributions allow for sampling to encode uncertainty stemming from HRUNet's limited performance. On this basis, we redefine HRUNet's output as a distribution of potential HR values, as opposed to the traditional emphasis on the single most probable HR value. The underlying goal is to discover the uncertainty arising from inherent noise in the input video. HRUNet is evaluated across 1, 098 videos from seven datasets, spanning three scenarios: undisturbed, motion-disturbed, and light-disturbed. The ensuing test outcomes demonstrate that uncertainty in the HR measurements increases significantly in the scenarios marked by disturbances, compared to that in the undisturbed scenario. Moreover, HRUNet outperforms state-of-the-art methods in HR accuracy when excluding HR values with $>$ 0. 4 uncertainty. This underscores that uncertainty emerges as an informative indicator of potentially erroneous HR measurements. With enhanced reliability affirmed, the VPPG technique holds the promise for applications in safety-critical domains.

Details DOI

ECAI Conference 2024 Conference Paper

Unified Video and Image Representation for Boosted Video Face Forgery Detection

Haotian Liu
Chenhui Pan
Yang Liu 0182
Guoying Zhao 0001
Xiaobai Li

Face forgery detection is crucial in preserving the security and integrity of facial data amidst the rapid developments in face manipulation techniques and deep generative models. Existing methods for video face forgery detection typically assume that all frames in a forged video are manipulated, while identifying partially forged videos with only a subset of altered frames is still a challenge to be solved. To address this issue, we propose a novel framework, i. e. , the UVIF, that utilizes additional annotated images to provide fine-grained supervision for detecting partial forgeries in videos. The UVIF integrates a unified encoder and a multi-task learning paradigm to model both facial videos and images for boosted video face forgery detection. A 2D backbone with temporal fusion modules is employed for the unified encoder. A pseudo labeling process is also designed for facial video frames to bridge the representation of individual video frames and static images. Extensive experiments on benchmark datasets demonstrate the effectiveness of our framework, outperforming state-of-the-art methods in detecting partially forged videos while introducing no additional computational overhead. Our code is available at https: //github. com/haotianll/UVIF.

Details

JBHI Journal 2022 Journal Article

Non-Contact Atrial Fibrillation Detection From Face Videos by Learning Systolic Peaks

Zhaodong Sun
Juhani Junttila
Mikko Tulppo
Tapio Seppanen
Xiaobai Li

Objective: We propose a non-contact approach for atrial fibrillation (AF) detection from face videos. Methods: Face videos, electrocardiography (ECG), and contact photoplethysmography (PPG) from 100 healthy subjects and 100 AF patients are recorded. Data recordings from healthy subjects are all labeled as healthy. Two cardiologists evaluated ECG recordings of patients and labeled each recording as AF, sinus rhythm (SR), or atrial flutter (AFL). We use the 3D convolutional neural network for remote PPG monitoring and propose a novel loss function (Wasserstein distance) to use the timing of systolic peaks from contact PPG as the label for our model training. Then a set of heart rate variability (HRV) features are calculated from the inter-beat intervals, and a support vector machine (SVM) classifier is trained with HRV features. Results: Our proposed method can accurately extract systolic peaks from face videos for AF detection. The proposed method is trained with subject-independent 10-fold cross-validation with 30 s video clips and tested on two tasks. 1) Classification of healthy versus AF: the accuracy, sensitivity, and specificity are 96. 00%, 95. 36%, and 96. 12%. 2) Classification of SR versus AF: the accuracy, sensitivity, and specificity are 95. 23%, 98. 53%, and 91. 12%. In addition, we also demonstrate the feasibility of non-contact AFL detection. Conclusion: We achieve good performance of non-contact AF detection by learning systolic peaks. Significance: non-contact AF detection can be used for self-screening of AF symptoms for suspectable populations at home or self-monitoring of AF recurrence after treatment for chronic patients.

Details DOI

IJCAI Conference 2021 Conference Paper

Dual-Cross Central Difference Network for Face Anti-Spoofing

Zitong Yu
Yunxiao Qin
Hengshuang Zhao
Xiaobai Li
Guoying Zhao

Face anti-spoofing (FAS) plays a vital role in securing face recognition systems. Recently, central difference convolution (CDC) has shown its excellent representation capacity for the FAS task via leveraging local gradient features. However, aggregating central difference clues from all neighbors/directions simultaneously makes the CDC redundant and sub-optimized in the training phase. In this paper, we propose two Cross Central Difference Convolutions (C-CDC), which exploit the difference of the center and surround sparse local features from the horizontal/vertical and diagonal directions, respectively. It is interesting to find that, with only five ninth parameters and less computational cost, C-CDC even outperforms the full directional CDC. Based on these two decoupled C-CDC, a powerful Dual-Cross Central Difference Network (DC-CDN) is established with Cross Feature Interaction Modules (CFIM) for mutual relation mining and local detailed representation enhancement. Furthermore, a novel Patch Exchange (PE) augmentation strategy for FAS is proposed via simply exchanging the face patches as well as their dense labels from random samples. Thus, the augmented samples contain richer live/spoof patterns and diverse domain distributions, which benefits the intrinsic and robust feature learning. Comprehensive experiments are performed on four benchmark datasets with three testing protocols to demonstrate our state-of-the-art performance.

PDF Details DOI