# SpinML: Customized Synthetic Data Generation for Private Training of Specialized ML Models

arXiv: https://arxiv.org/abs/2503.03160 · HTML: https://mlsanity.com/paper/2503.03160/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 68 references that are in the map.
- First posted on arXiv 5 Mar 2025.

## Builds on (most cited first, 25 of 25)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Deep Learning with Differential Privacy](https://mlsanity.com/paper/1607.00133/) (2016)
- [Inverting Gradients -- How easy is it to break privacy in federated learning?](https://mlsanity.com/paper/2003.14053/) (2020)
- [Federated Learning for Mobile Keyboard Prediction](https://mlsanity.com/paper/1811.03604/) (2018)
- [Synthetic Data from Diffusion Models Improves ImageNet Classification](https://mlsanity.com/paper/2304.08466/) (2023)
- [See through Gradients: Image Batch Recovery via GradInversion](https://mlsanity.com/paper/2104.07586/) (2021)
- [Streaming End-to-end Speech Recognition For Mobile Devices](https://mlsanity.com/paper/1811.06621/) (2018)
- [Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge Computing](https://mlsanity.com/paper/1905.10083/) (2019)
- [Fake It Till You Make It: Face analysis in the wild using synthetic data alone](https://mlsanity.com/paper/2109.15102/) (2021)
- [LDP-Fed: Federated Learning with Local Differential Privacy](https://mlsanity.com/paper/2006.03637/) (2020)
- [Driving Policy Transfer via Modularity and Abstraction](https://mlsanity.com/paper/1804.09364/) (2018)
- [Turbo-Aggregate: Breaking the Quadratic Aggregation Barrier in Secure Federated Learning](https://mlsanity.com/paper/2002.04156/) (2020)
- [Training on Thin Air: Improve Image Classification with Generated Data](https://mlsanity.com/paper/2305.15316/) (2023)
- [Machine learning for Internet of Things data analysis: A survey](https://mlsanity.com/paper/1802.06305/) (2018)
- [Context-Aware Generative Adversarial Privacy](https://mlsanity.com/paper/1710.09549/) (2017)
- [DP-Image: Differential Privacy for Image Data in Feature Space](https://mlsanity.com/paper/2103.07073/) (2021)
- [Fine-grained Controllable Video Generation via Object Appearance and Context](https://mlsanity.com/paper/2312.02919/) (2023)
- [GANs for Biological Image Synthesis](https://mlsanity.com/paper/1708.04692/) (2017)
- [Learning to Protect Communications with Adversarial Neural Cryptography](https://mlsanity.com/paper/1610.06918/) (2016)
- [Preserving Privacy in Personalized Models for Distributed Mobile Services](https://mlsanity.com/paper/2101.05855/) (2021)
- [SkillVet: Automated Traceability Analysis of Amazon Alexa Skills](https://mlsanity.com/paper/2103.02637/) (2021)
- [Privacy-Utility Trades in Crowdsourced Signal Map Obfuscation](https://mlsanity.com/paper/2201.04782/) (2022)

## Abstract (arXiv)

Specialized machine learning (ML) models tailored to users needs and requests are increasingly being deployed on smart devices with cameras, to provide personalized intelligent services taking advantage of camera data. However, two primary challenges hinder the training of such models: the lack of publicly available labeled data suitable for specialized tasks and the inaccessibility of labeled private data due to concerns about user privacy. To address these challenges, we propose a novel system SpinML, where the server generates customized Synthetic image data to Privately traIN a specialized ML model tailored to the user request, with the usage of only a few sanitized reference images from the user. SpinML offers users fine-grained, object-level control over the reference images, which allows user to trade between the privacy and utility of the generated synthetic data according to their privacy preferences. Through experiments on three specialized model training tasks, we demonstrate that our proposed system can enhance the performance of specialized models without compromising users privacy preferences.