KDPhys: An Attention Guided 3D to 2D Knowledge Distillation for Real-time Video-Based Physiological Measurement
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917180472295424 |
|---|---|
| author | Sahoo, Nicky Nirlipta Sachidanand, VS Gayathri, Matcha Naga Murugesan, Balamurali Ram, Keerthi Joseph, Jayaraj Sivaprakasam, Mohanasankar |
| author_facet | Sahoo, Nicky Nirlipta Sachidanand, VS Gayathri, Matcha Naga Murugesan, Balamurali Ram, Keerthi Joseph, Jayaraj Sivaprakasam, Mohanasankar |
| contents | Camera-based physiological monitoring, such as remote photoplethysmography (rPPG), captures subtle variations in skin optical properties caused by pulsatile blood volume changes using standard digital camera sensors. The demand for real-time, non-contact physiological measurement has increased significantly, particularly during the SARS-CoV-2 pandemic, to support telehealth and remote health monitoring applications. In this work, we propose an attention-based knowledge distillation (KD) framework, termed KDPhys, for extracting rPPG signals from facial video sequences. The proposed method distills global temporal representations from a 3D convolutional neural network (CNN) teacher model to a lightweight 2D CNN student model through effective 3D-to-2D feature distillation. To the best of our knowledge, this is the first application of knowledge distillation in the rPPG domain. Furthermore, we introduce a Distortion Loss incorporating Shape and Time (DILATE), which jointly accounts for both morphological and temporal characteristics of rPPG signals. Extensive qualitative and quantitative evaluations are conducted on three benchmark datasets. The proposed model achieves a significant reduction in computational complexity, using only half the parameters of existing methods while operating 56.67% faster. With just 0.23M parameters, it achieves an 18.15% reduction in Mean Absolute Error (MAE) compared to state-of-the-art approaches, attaining an average MAE of 1.78 bpm across all datasets. Additional experiments under diverse environmental conditions and activity scenarios further demonstrate the robustness and adaptability of the proposed approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_00714 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | KDPhys: An Attention Guided 3D to 2D Knowledge Distillation for Real-time Video-Based Physiological Measurement Sahoo, Nicky Nirlipta Sachidanand, VS Gayathri, Matcha Naga Murugesan, Balamurali Ram, Keerthi Joseph, Jayaraj Sivaprakasam, Mohanasankar Image and Video Processing Camera-based physiological monitoring, such as remote photoplethysmography (rPPG), captures subtle variations in skin optical properties caused by pulsatile blood volume changes using standard digital camera sensors. The demand for real-time, non-contact physiological measurement has increased significantly, particularly during the SARS-CoV-2 pandemic, to support telehealth and remote health monitoring applications. In this work, we propose an attention-based knowledge distillation (KD) framework, termed KDPhys, for extracting rPPG signals from facial video sequences. The proposed method distills global temporal representations from a 3D convolutional neural network (CNN) teacher model to a lightweight 2D CNN student model through effective 3D-to-2D feature distillation. To the best of our knowledge, this is the first application of knowledge distillation in the rPPG domain. Furthermore, we introduce a Distortion Loss incorporating Shape and Time (DILATE), which jointly accounts for both morphological and temporal characteristics of rPPG signals. Extensive qualitative and quantitative evaluations are conducted on three benchmark datasets. The proposed model achieves a significant reduction in computational complexity, using only half the parameters of existing methods while operating 56.67% faster. With just 0.23M parameters, it achieves an 18.15% reduction in Mean Absolute Error (MAE) compared to state-of-the-art approaches, attaining an average MAE of 1.78 bpm across all datasets. Additional experiments under diverse environmental conditions and activity scenarios further demonstrate the robustness and adaptability of the proposed approach. |
| title | KDPhys: An Attention Guided 3D to 2D Knowledge Distillation for Real-time Video-Based Physiological Measurement |
| topic | Image and Video Processing |
| url | https://arxiv.org/abs/2601.00714 |