# Universal Embedding Function for Traffic Classification via QUIC Domain Recognition Pretraining: A Transfer Learning Success

arXiv: https://arxiv.org/abs/2502.12930 · HTML: https://mlsanity.com/paper/2502.12930/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 54 references that are in the map.
- First posted on arXiv 18 Feb 2025.

## Builds on (most cited first, 23 of 23)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [Supervised Contrastive Learning](https://mlsanity.com/paper/2004.11362/) (2020)
- [Billion-scale similarity search with GPUs](https://mlsanity.com/paper/1702.08734/) (2017)
- [CosFace: Large Margin Cosine Loss for Deep Face Recognition](https://mlsanity.com/paper/1801.09414/) (2018)
- [Fine-tuning CNN Image Retrieval with No Human Annotation](https://mlsanity.com/paper/1711.02512/) (2017)
- [Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?](https://mlsanity.com/paper/2003.11539/) (2020)
- [Explicit Inductive Bias for Transfer Learning with Convolutional Networks](https://mlsanity.com/paper/1802.01483/) (2018)
- [On Embeddings for Numerical Features in Tabular Deep Learning](https://mlsanity.com/paper/2203.05556/) (2022)
- [Spreading vectors for similarity search](https://mlsanity.com/paper/1806.03198/) (2018)
- [Fine-tuning can cripple your foundation model; preserving features may be the solution](https://mlsanity.com/paper/2308.13320/) (2023)
- [ET-BERT: A Contextualized Datagram Representation with Pre-training Transformers for Encrypted Traffic Classification](https://mlsanity.com/paper/2202.06335/) (2022)
- [CLIP Itself is a Strong Fine-tuner: Achieving 85.7% and 88.0% Top-1 Accuracy with ViT-B and ViT-L on ImageNet](https://mlsanity.com/paper/2212.06138/) (2022)
- [How to Achieve High Classification Accuracy with Just a Few Labels: A Semi-supervised Approach Using Sampled Packets](https://mlsanity.com/paper/1812.09761/) (2018)
- [Deep Learning and Traffic Classification: Lessons learned from a commercial-grade dataset with hundreds of encrypted and zero-day applications](https://mlsanity.com/paper/2104.03182/) (2021)
- [Fine-grained TLS services classification with reject option](https://mlsanity.com/paper/2202.11984/) (2022)
- [Multitask Learning for Network Traffic Classification](https://mlsanity.com/paper/1906.05248/) (2019)
- [Google Landmark Recognition 2020 Competition Third Place Solution](https://mlsanity.com/paper/2010.05350/) (2020)
- [A Lightweight, Efficient and Explainable-by-Design Convolutional Neural Network for Internet Traffic Classification](https://mlsanity.com/paper/2202.05535/) (2022)
- [Replication: Contrastive Learning and Data Augmentation in Traffic Classification Using a Flowpic Input Representation](https://mlsanity.com/paper/2309.09733/) (2023)
- [Many or Few Samples? Comparing Transfer, Contrastive and Meta-Learning in Encrypted Traffic Classification](https://mlsanity.com/paper/2305.12432/) (2023)
- [Data Augmentation for Traffic Classification](https://mlsanity.com/paper/2401.10754/) (2024)
- [DataZoo: Streamlining Traffic Classification Experiments](https://mlsanity.com/paper/2310.19568/) (2023)

## Built on by (most cited first, 1 of 1)

- [When Simple Model Just Works: Is Network Traffic Classification in Crisis?](https://mlsanity.com/paper/2506.08655/) (2025)

## Abstract (arXiv)

Encrypted traffic classification (TC) methods must adapt to new protocols and extensions as well as to advancements in other machine learning fields. In this paper, we adopt a transfer learning setup best known from computer vision. We first pretrain an embedding model on a complex task with a large number of classes and then transfer it to seven established TC datasets. The pretraining task is recognition of SNI domains in encrypted QUIC traffic, which in itself is a challenge for network monitoring due to the growing adoption of TLS Encrypted Client Hello. Our training pipeline -- featuring a disjoint class setup, ArcFace loss function, and a modern deep learning architecture -- aims to produce universal embeddings applicable across tasks. A transfer method based on model fine-tuning surpassed SOTA performance on nine of ten downstream TC tasks, with an average improvement of 6.4%. Furthermore, a comparison with a baseline method using raw packet sequences revealed unexpected findings with potential implications for the broader TC field. We released the model architecture, trained weights, and codebase for transfer learning experiments.