惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

爱范儿
爱范儿
WordPress大学
WordPress大学
博客园 - 【当耐特】
The Cloudflare Blog
B
Blog
Last Week in AI
Last Week in AI
小众软件
小众软件
量子位
S
SegmentFault 最新的问题
V
Visual Studio Blog
博客园 - 叶小钗
美团技术团队
阮一峰的网络日志
阮一峰的网络日志
Hugging Face - Blog
Hugging Face - Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
宝玉的分享
宝玉的分享
A
About on SuperTechFans
雷峰网
雷峰网
J
Java Code Geeks
Microsoft Azure Blog
Microsoft Azure Blog
腾讯CDC
MongoDB | Blog
MongoDB | Blog
酷 壳 – CoolShell
酷 壳 – CoolShell
Martin Fowler
Martin Fowler

eess.SP updates on arXiv.org

ECG-biometrics-bench: A Unified Framework for Reproducible Benchmarking of ECG Biometrics Physiology-Aware Masked Cross-Modal Reconstruction for Biosignal Representation Learning Towards Improving Speaker Distance Estimation through Generative Impulse Response Augmentation Federated Learning with Hypergradient-based Online Update of Aggregation Weights Soft Graph Diffusion Transformer for MIMO Detection SPLICE: Latent Diffusion over JEPA Embeddings for Conformal Time-Series Inpainting Sequential Inference for Gaussian Processes: A Signal Processing Perspective Statistical Channel Fingerprint Construction for Massive MIMO: A Unified Tensor Learning Framework Recent Advances in mm-Wave and Sub-THz/THz Oscillators for FutureG Technologies Cross-Subject Generalization for EEG Decoding: A Survey of Deep Learning Methods Super-resolution Multi-signal Direction-of-Arrival Estimation by Hankel-structured Sensing and Decomposition Hankel and Toeplitz Rank-1 Decomposition of Arbitrary Matrices with Applications to Signal Direction-of-Arrival Estimation Adaptive Transform Coding for Semantic Compression EdgeSpike: Spiking Neural Networks for Low-Power Autonomous Sensing in Edge IoT Architectures Sparse Graph Learning from Sparse Data via Fiedler Number Maximization A Deep Learning Model for Battery State Prediction towards Intelligent Energy Management Transfer Learning for Tonal Noise Prediction in VRF Units Using Thermodynamic and Vibration Signals EVT-Based Generative AI for Tail-Aware Channel Estimation Monitoring exposure-length variations in submarine power cables using distributed fiber-optic sensing BandRouteNet: An Adaptive Band Routing Neural Network for EEG Artifact Removal Phase-Separated Complex Hilbert PCA on Markerless 3D Pose Estimation Data: A Global Phase Network and Its Extension to a Continuous Field on the Body Surface Selective Correlation Based Knowledge Distillation for Ground Reaction Force Estimation Deep Learning-Enabled Dissolved Oxygen Sensing in Biofouling Environments for Ocean Monitoring Speech Enhancement Based on Drifting Models Robust and Clinically Reliable EEG Biomarkers: A Cross Population Framework for Generalizable Parkinson's Disease Detection An AI-Based Supervisory Measurement Integrity Validation Layer for Cyber-Resilient AC/DC Protection in Inverter-Based Microgrids Explainable AI in Speaker Recognition -- Making Latent Representations Understandable Time-Localized Parametric Decomposition of Respiratory Airflow for Sub-Breath Analysis NAKUL-Med: Spectral-Graph State Space Models with Dynamics Kernels for Medical Signals An Algorithm for On-Sensor Agnostic Detection of Changes in Human Activity for Ultra-Low-Power Applications
Tokenization vs. Augmentation: A Systematic Study of Writ...
Jindong Li, Dario Zanca, Vincent Christlein, Tim Hamann, Jens Ba · 2026-02-25 · via eess.SP updates on arXiv.org

Inertial measurement unit-based online handwriting recognition enables the recognition of input signals collected across different writing surfaces but remains challenged by uneven character distributions and inter-writer variability. In this work, we systematically investigate two strategies to address these issues: sub-word tokenization and concatenation-based data augmentation. Our experiments on the OnHW-Words500 dataset reveal a clear dichotomy between handling inter-writer and intra-writer variance. On the writer-independent split, structural abstraction via Bigram tokenization significantly improves performance to unseen writing styles, reducing the word error rate (WER) from 15.40% to 12.99%. In contrast, on the writer-dependent split, tokenization degrades performance due to vocabulary distribution shifts between the training and validation sets. Instead, our proposed concatenation-based data augmentation acts as a powerful regularizer, reducing the character error rate by 34.5% and the WER by 25.4%. Further analysis shows that short, low-level tokens benefit model performance and that concatenation-based data augmentation performance gain surpasses those achieved by proportionally extended training. These findings reveal a clear variance-dependent effect: sub-word tokenization primarily mitigates inter-writer stylistic variability, whereas concatenation-based data augmentation effectively compensates for intra-writer distributional sparsity.