Stabilizing Temporal Inference Dynamics for Online Surgical Phase Recognition
Quick Take
A framework stabilizes temporal inference in online surgical phase recognition, enhancing reliability and reducing prediction fragmentation.
Key Points
- Introduces a unified Train-Inference-Evaluation framework.
- Utilizes Temporal Error-Cascade loss for training stability.
- Implements Evidence-Gated Transition Predictor for inference.
📖 Reader Mode
~2 min readAbstract:Online Surgical Phase Recognition (SPR) models can reach high frame-wise accuracy, yet their predictions often lack temporal stability, fragmenting workflow understanding and reducing the reliability of downstream assistance. We show that this instability is not random noise but arises from two mechanisms: early misclassifications corrupt temporal feature states and propagate forward to form error cascades, and phase transitions follow evidence-accumulation dynamics whereas most online SPR systems rely on memoryless frame-wise decisions, making them sensitive to transient confidence fluctuations. We propose a unified Train-Inference-Evaluation framework that explicitly stabilizes temporal inference dynamics using model-agnostic, plug-and-play components. For training, the Temporal Error-Cascade (TEC) loss suppresses error onset and mitigates forward error propagation by stabilizing temporal feature evolution. For inference, the Evidence-Gated Transition Predictor (EGTP) enforces evidence-driven state transitions, allowing phase changes only when accumulated evidence exceeds a confidence boundary. For evaluation, we introduce the Temporal Fragmentation Index (TFI), a reliability-aware metric that quantifies instability-induced temporal disagreement beyond conventional frame-wise and token-based measures. Experiments on Cholec80 and AutoLaparo across three representative backbones show that the proposed framework substantially improves temporal stability and reduces prediction fragmentation, while maintaining or modestly improving frame-wise performance.
| Comments: | Early accepted by MICCAI 2026 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.16387 [cs.CV] |
| (or arXiv:2605.16387v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.16387 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yang Liu [view email]
[v1]
Mon, 11 May 2026 16:56:15 UTC (2,457 KB)
— Originally published at arxiv.org
More from arXiv cs.CV
See more →GeoSym127K: Scalable Symbolically-verifiable Synthesis for Multimodal Geometric Reasoning
GeoSym127K introduces a scalable neuro-symbolic framework for enhanced geometric reasoning in multimodal models.