Explaining RhythmFormer: A Systematic XAI Analysis of Periodic Sparse Attention for Remote Photoplethysmography

• URL: http://arxiv.org/abs/2606.13839v1
• Date: Thu, 11 Jun 2026 19:10:06 GMT
• Title: Explaining RhythmFormer: A Systematic XAI Analysis of Periodic Sparse Attention for Remote Photoplethysmography
• Abstract summary: Photoplethys transformers achieve low heart-rate error on benchmarks, yet their decisions remain opaque.<n>Existing XAI is dominated by qualitative heatmap inspection without quantitative faithfulness metrics or physiology-grounded validation.<n>We adapt four methods (raw attention, rollout, flow, Beyond Intuition) to quantifying RhythmFormer's bi-level attention with top-$k$ selection.
• Score: 1.647210198730602
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Remote photoplethysmography (rPPG) transformers achieve low heart-rate error on benchmarks, yet their decisions remain opaque--a growing concern as rPPG moves toward clinical heart rate estimation. Existing rPPG XAI is dominated by qualitative heatmap inspection without quantitative faithfulness metrics or physiology-grounded validation, leaving a gap between visual plausibility and auditable evidence. We address this gap. First, we adapt four attribution methods (raw attention, rollout, flow, Beyond Intuition) to RhythmFormer's bi-level routing attention with top-$k$ selection. Second, we introduce a skin coverage metric quantifying how much attribution mass falls on skin regions. Third, we adapt the SaCo faithfulness coefficient from its original classification setting to rPPG regression by using the MAE between original and perturbed predicted rPPG waveforms as the perturbation impact. Applying these tools, we quantify a multi-hop leakage effect under sparse top-$k$ routing: attention rollout and flow almost completely restores the connections that individual refined-attention layers explicitly set to zero. Beyond Intuition mitigates this via its value-projection-weighted rollout and gradient-supported mask, attaining the highest median refined skin coverage ($0.83$ vs. $0.57$ for vanilla rollout) and faithfulness ($F=0.92$) among the evaluated methods on UBFC-rPPG. Validation across diverse datasets and model variants is needed. A case study on a low-SaCo outlier further shows all four methods recovering consistently once an artefactual region is replaced, suggesting consistent SaCo behavior across attribution families in this illustrative case. Together, these metrics move XAI for rPPG toward auditable numerical evidence about spatial alignment and perturbation faithfulness, i.e. trustworthy rPPG XAI.

Related papers

• Not All Relations Are Equal: Relation-Balanced and Calibrated Graph Learning for Provenance-Based Intrusion Detection [66.44527094471619]
  We present RECAL, an unsupervised framework using relation-balanced masked graph learning to better capture rare interaction patterns.<n>On three DARPA E3 datasets, RECAL achieves F1 scores of 99.99%, 99.93%, and 99.99%, outperforming the best baseline on each dataset by 0.88, 0.82, and 0.42 percentage points, respectively.
  arXiv  Detail & Related papers  (2026-09-15T00:31:01Z)
• Beyond Ambiguous Visual Cues: Studying Physiological Disruptions and Cross-Modal Inconsistencies in Deepfake Videos [7.77578121448493]
  Current detectors underexplore the cross-level relationship between facial features and physiological dynamics, often relying on late fusion or r features alone.<n>We propose a bidirectional co-attention fusion detector that captures the cross-level dependencies between pulse dynamics and facial motion to learn a robust, joint authenticity representation.
  arXiv  Detail & Related papers  (2026-09-11T10:12:48Z)
• ToxLens: A Reproducible Graph-Learning Framework for Leakage-Aware, Uncertainty-Calibrated Molecular Toxicity Prediction [0.0]
  ToxLens is a graph-learning framework for 11 toxicity endpoints spanning Ames mutagenicity, acute oral toxicity, hERG inhibition, and Tox21 nuclear-receptor and stress-response assays.<n>The workflow combines conservative chemical curation, sphere-exclusion filtering, a leakage-aware UMAP-HDBSCAN split, parallel graph and global-feature encoders joined by late concatenation.<n>On the leakage-controlled test fold, a five-seed soft-voting ensemble achieved a Matthews correlation coefficient score of 0.44, an area under the receiver operating characteristic curve score of 0.83, and an area under the
  arXiv  Detail & Related papers  (2026-08-31T08:58:51Z)
• Gated Spatial Redundancy Projection for Pathology Transformer Attentions [9.268049949671958]
  Gated Spatial Redundancy Projection (Gated SRP) is a lightweight correction module for self-attention layers.<n>Gated SRP estimates a local redundancy axis from neighbouring value, projects the attention output onto this axis, and applies a learned signed gate to correct the redundancy-aligned component geometrically.
  arXiv  Detail & Related papers  (2026-08-08T23:59:10Z)
• CoSA: Correlation-Guided Change Attention with Learnable Residual Gating for Remote Sensing Change Detection [1.9364098361060789]
  We propose Context Sampling Attention (CoSA), a lightweight decoder-side refinement module.<n>CoSA explicitly leverages bi-temporal feature correlation as a control signal for adaptive change-aware feature enhancement.<n>Experiments on four benchmark datasets demonstrate consistent improvements over strong baselines.
  arXiv  Detail & Related papers  (2026-06-20T07:56:09Z)
• LQ-rPPG: A Label-Quantized Coarse-to-Fine Learning Framework for Remote Physiological Measurement [9.606144267032215]
  LQ-r is a label-to-fine learning framework for robust r estimation.<n>It transforms continuous signals into multi-bit quantized pseudo labels with reduced noise and variability.<n>LQ-r achieves robust and general r estimation even under challenging conditions.
  arXiv  Detail & Related papers  (2026-05-22T02:46:11Z)
• MoASE++: Mixture of Activation Sparsity Experts with Domain-Adaptive On-policy Distillation for Continual Test Time Adaptation [65.068801413044]
  Continual test-time adaptation adapts a source-pretrained model to non-stationary, unlabeled target streams.<n>We introduce MoASE, a mixture-of-experts that disentangles domain-agnostic structure from domain-specific texture.
  arXiv  Detail & Related papers  (2026-05-18T01:52:12Z)
• Posterior Augmented Flow Matching [64.1559809786948]
  Posterior-Augmented Flow Matching (PAFM) is a generalization of flow matching (FM)<n>We show that PAFM improves over FM by up to 3.4 FID50K across different model scales.
  arXiv  Detail & Related papers  (2026-05-01T17:59:59Z)
• Intervention-Based Self-Supervised Learning: A Causal Probe Paradigm for Remote Photoplethysmography [38.62089038790769]
  We propose a new SSL paradigm, Causal Probing (PCP)<n>PCP treats the latent r signal as the underlying physical source and the resulting pixel chrominance variations as its visual manifestation.<n>Our editor achieves both in-domain and cross-domain performance on challenging datasets such as VIPL-HR and MMPD-HR.
  arXiv  Detail & Related papers  (2026-04-26T09:56:09Z)
• BPC-Net: Annotation-Free Skin Lesion Segmentation via Boundary Probability Calibration [15.992042021607643]
  We propose BPC-Net, a boundary probability calibration framework for annotation-free skin lesion segmentation.<n>The proposed framework achieves state-of-the-art performance among published unsupervised methods, reaching a macro-average Dice coefficient and Jaccard index of 85.80% and 76.97%, respectively.
  arXiv  Detail & Related papers  (2026-04-07T08:43:34Z)
• ConsistentRFT: Reducing Visual Hallucinations in Flow-based Reinforcement Fine-Tuning [85.20505958752928]
  Reinforcement Fine-Tuning (RFT) on flow-based models is crucial for preference alignment.<n>RFT often introduce visual hallucinations like over-optimized details and semantic misalignment.<n>This work preliminarily explores why visual hallucinations arise and how to reduce them.
  arXiv  Detail & Related papers  (2026-02-03T11:49:46Z)
• Continual Action Quality Assessment via Adaptive Manifold-Aligned Graph Regularization [53.82400605816587]
  Action Quality Assessment (AQA) quantifies human actions in videos, supporting applications in sports scoring, rehabilitation, and skill evaluation.<n>A major challenge lies in the non-stationary nature of quality distributions in real-world scenarios.<n>We introduce Continual AQA (CAQA), which equips with Continual Learning capabilities to handle evolving distributions.
  arXiv  Detail & Related papers  (2025-10-08T10:09:47Z)
• SQUWA: Signal Quality Aware DNN Architecture for Enhanced Accuracy in Atrial Fibrillation Detection from Noisy PPG Signals [37.788535094404644]
  Atrial fibrillation (AF) significantly increases the risk of stroke, heart disease, and mortality. Photoplethysmography ( PPG) signals are susceptible to corruption from motion artifacts and other factors often encountered in ambulatory settings. We propose a novel deep learning model, designed to learn how to retain accurate predictions from partially corrupted PPG.
  arXiv  Detail & Related papers  (2024-04-15T01:07:08Z)
• Dense Label Encoding for Boundary Discontinuity Free Rotation Detection [69.75559390700887]
  This paper explores a relatively less-studied methodology based on classification. We propose new techniques to push its frontier in two aspects. Experiments and visual analysis on large-scale public datasets for aerial images show the effectiveness of our approach.
  arXiv  Detail & Related papers  (2020-11-19T05:42:02Z)
• Meta-rPPG: Remote Heart Rate Estimation Using a Transductive Meta-Learner [15.701494085639007]
  Remote heart rate estimation is done using remote photoplethysmography (r) We propose a transductive meta-learner that takes unlabeled samples during testing (deployment) for a self-supervised weight adjustment.
  arXiv  Detail & Related papers  (2020-07-14T03:01:46Z)
• Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection [85.53263670166304]
  One-stage detector basically formulates object detection as dense classification and localization. Recent trend for one-stage detectors is to introduce an individual prediction branch to estimate the quality of localization. This paper delves into the representations of the above three fundamental elements: quality estimation, classification and localization.
  arXiv  Detail & Related papers  (2020-06-08T07:24:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.