Research map / Label-efficient and robust learning / Recognition and long-tail
Yet Another Traffic Classifier: A Masked Autoencoder Based Traffic Transformer with Multi-Level Flow Representation
Ruijie Zhao, Mingwei Zhan, Xianwen Deng, Yanhao Wang, Yijun Wang, Guan Gui, Zhi Xue
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use YaTC instead of the strongest reported baseline | traffic classification · ISCXVPN2016 | F1: 88.61% -> 98.04% | Improved |
| Use YaTC instead of the strongest reported baseline | traffic classification · ISCXTor2016 | F1: 79.99% -> 99.72% | Improved |
| Use YaTC instead of the strongest reported baseline | traffic classification · USTC-TFC2016 | F1: 96.95% -> 97.86% | Improved |
| Use YaTC instead of the strongest reported baseline | traffic classification · CICIoT2022 | F1: 90.49% -> 96.58% | Improved |
| Add packet-level and flow-level attention rather than global attention (ablation) | YaTC · four reported datasets | F1 | Mixed |
| Add packet-level attention (ablation) | YaTC · four reported datasets | F1 | Improved |
| Add flow-level attention (ablation) | YaTC · four reported datasets | F1 | Improved |
| Add flow-level stacking to MFR (ablation) | YaTC · four reported datasets | F1 | Improved |
| Share parameters between the two traffic encoders (ablation) | YaTC fine-tuning · four reported datasets | F1 | Improved |
| Add pre-training (ablation) | YaTC · four reported datasets | F1 | Improved |
| Use MFR rather than directly transforming raw traffic bytes (ablation) | YaTC without pre-training · four reported datasets | F1 | Improved |
| Use a 90% masking ratio in MAE pre-training (ablation) | YaTC · most datasets | F1 | Improved |
Closest papers by meaning
- MIETT: Multi-Instance Encrypted Traffic Transformer for Encrypted Traffic Classification (AAAI 2025 · similarity 0.85)
- Revolutionizing Encrypted Traffic Classification with MH-Net: A Multi-View Heterogeneous Graph Model (AAAI 2025 · similarity 0.84)
- Dual-Channel Interactive Graph Transformer for Traffic Classification with Message-Aware Flow Representation (AAAI 2025 · similarity 0.84)
- SSL-STMFormer Self-Supervised Learning Spatio-Temporal Entanglement Transformer for Traffic Flow Prediction (AAAI 2025 · similarity 0.83)
- MalDetectFormer: Leveraging Sparse SpatioTemporal Information for Effective Malicious Traffic Detection (AAAI 2025 · similarity 0.82)
- WardropNet: Traffic Flow Predictions via Equilibrium-Augmented Learning (ICLR 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
