arXiv:2606. 15377v1 Announce Type: cross Abstract: Inaccurately labeled training data, or "label noise", poses a significant threat to the integrity of supervised machine learning models.
Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.
The paper examines how varying input noise characteristics—type, scale, and complexity—affect neural network robustness in geophysical tasks such as first break picking and denoising. By training models on fixed noise settings and testing them on both seen and unseen noise scenarios, the study constructs a robustness matrix that reveals how larger noise scales improve generalization and how aligning noise type with task complexity and architecture maximizes performance. Training with compound noise mixtures further mitigates weaknesses of single-noise training, acting as an implicit regularizer that enhances robustness under out‑of‑distribution conditions.
The paper presents a framework that adapts general-purpose Vision Foundation Models (VFMs) to seismic data denoising using Parameter‑Efficient Fine‑Tuning with Low‑Rank Adaptation (LoRA). It introduces a kurtosis‑guided unsupervised test‑time adaptation module that updates only LoRA parameters to self‑calibrate for site‑specific noise without ground truth. Experiments on exploration seismic images and DAS data demonstrate that the approach matches or surpasses domain‑specific models and generalizes well to unseen cross‑site data.
The paper investigates whether seismic foundation models can transcend being mere feature extractors by training a Mamba encoder‑decoder on 3 million waveform windows with varied self‑supervised objectives, quantization bottlenecks, and mixers. Experiments show that changing the training target redistributes performance across downstream tasks rather than improving it overall; adding masking improves few‑shot P‑wave picking but harms classification accuracy. The authors conclude that current models act as feature extractors, and that a learning target spanning seismic events—capturing the stochastic inter‑event timescale—is needed to achieve true foundation‑model capabilities.
arXiv:2606. 10069v3 Announce Type: replace Abstract: In this paper we build upon a previous study in which we demonstrated, using XGBoost and earthquake catalogue data from Japan and Chile, that a set of 60 seismic statistical features (SSFs) had much greater predictive value than a set of 428 generic time series features from the tsfresh package.
The paper proposes a method that first pre‑trains a feature extractor on the target dataset using in‑domain self‑supervised learning (SSL) without labels, then performs standard supervised training on the same noisy dataset. This two‑stage approach eliminates the need for a clean label subset and consistently improves classification accuracy and label‑error detection across synthetic and real‑world noise, especially as noise rates increase. Experiments show that the method matches or surpasses ImageNet and DinoV2 pre‑training, particularly under high noise conditions.
arXiv:2606. 10069v1 Announce Type: new Abstract: In this paper we build upon a previous study in which we demonstrated, using XGBoost and earthquake catalogue data from Japan and Chile, that a set of 60 seismic statistical features (SSFs) had much greater predictive value than a set of 428 generic time series features from the tsfresh package.