mlsanityData to 14 Oct 2026

Qimera: Data-free Quantization with Synthetic Boundary Supporting Samples

arXiv 2111.02625 ↗First posted 4 Nov 2021cs.LG cs.AI cs.CVcited by 40

Not on the map. Why?

Abstract (arXiv)

Model quantization is known as a promising method to compress deep neural networks, especially for inferences on lightweight mobile or edge devices. However, model quantization usually requires access to the original training data to maintain the accuracy of the full-precision models, which is often infeasible in real-world scenarios for security and privacy issues. A popular approach to perform quantization without access to the original data is to use synthetically generated samples, based on batch-normalization statistics or adversarial learning. However, the drawback of such approaches is that they primarily rely on random noise input to the generator to attain diversity of the synthetic samples. We find that this is often insufficient to capture the distribution of the original data, especially around the decision boundaries. To this end, we propose Qimera, a method that uses superposed latent embeddings to generate synthetic boundary supporting samples. For the superposed embeddings to better reflect the original distribution, we also propose using an additional disentanglement mapping layer and extracting information from the full-precision model. The experimental results show that Qimera achieves state-of-the-art performances for various settings on data-free quantization. Code is available at https://github.com/iamkanghyunchoi/qimera.

Key facts. Cited by 40 papers in the map (data through 14 Oct 2026). Builds on 42 of its 54 references that are in the map. First posted on arXiv 4 Nov 2021.

Builds on

42 of 54 references.

References · 42

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  4. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  5. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  6. 2013Intriguing properties of neural networkscited by 5,155
  7. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  8. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,884
  9. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,329
  10. 2014Conditional Generative Adversarial Netscited by 2,162
  11. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,636
  12. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,593
30 more
  1. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,432
  2. 2017In-Datacenter Performance Analysis of a Tensor Processing Unitcited by 618
  3. 2016DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradientscited by 569
  4. 2016Conditional Image Synthesis With Auxiliary Classifier GANscited by 565
  5. 2019Dreaming to Distill: Data-free Knowledge Transfer via DeepInversioncited by 423
  6. 2019Q-BERT: Hessian Based Ultra Low Precision Quantization of BERTcited by 359
  7. 2019SinGAN: Learning a Generative Model from a Single Natural Imagecited by 348
  8. 2019Data-Free Quantization Through Weight Equalization and Bias Correctioncited by 336
  9. 2019Semi-supervised Domain Adaptation via Minimax Entropycited by 335
  10. 2020ZeroQ: A Novel Zero Shot Quantization Frameworkcited by 269
  11. 2019Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNetcited by 260
  12. 2019Data-Free Learning of Student Networkscited by 204
  13. 2017Latent Space Oddity: on the Curvature of Deep Generative Modelscited by 174
  14. 2020Universal Domain Adaptation through Self Supervisioncited by 172
  15. 2017Modulating early visual processing by languagecited by 167
  16. 2018MixUp as Locally Linear Out-Of-Manifold Regularizationcited by 158
  17. 2019Improving Neural Network Quantization without Retraining using Outlier Channel Splittingcited by 151
  18. 2017Data-Free Knowledge Distillation for Deep Neural Networkscited by 136
  19. 2019Zero-shot Knowledge Transfer via Adversarial Belief Matchingcited by 129
  20. 2019Zero-Shot Knowledge Distillation in Deep Networkscited by 121
  21. 2019Defensive Quantization: When Efficiency Meets Robustnesscited by 95
  22. 2019Data-Free Adversarial Distillationcited by 90
  23. 2020Generative Low-bitwidth Data Free Quantizationcited by 88
  24. 2020Data-Free Network Quantization With Adversarial Knowledge Distillationcited by 80
  25. 2017Towards Effective Low-bitwidth Convolutional Neural Networkscited by 78
  26. 2021Diversifying Sample Generation for Accurate Data-Free Quantizationcited by 63
  27. 2018Knowledge Distillation with Adversarial Samples Supporting Decision Boundarycited by 49
  28. 2021Zero-shot Adversarial Quantizationcited by 44
  29. 2019Characterizing the Decision Boundary of Deep Neural Networkscited by 32
  30. 2021AutoReCon: Neural Architecture Search-based Reconstruction for Data-free Compressioncited by 17

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 1; 2022: 10; 2023: 10; 2024: 8; 2025: 7; 2026: 4.

40 papers.

Sort:

By area:

  1. 2021IntraQ: Learning Synthetic Images with Intra-Class Heterogeneity for Zero-Shot Network Quantizationcited by 43
  2. 2022PSAQ-ViT V2: Towards Accurate and General Data-Free Quantization for Vision Transformerscited by 42
  3. 2023Adaptive Data-Free Quantizationcited by 26
  4. 2023Unified Data-Free Compression: Pruning and Quantization without Fine-Tuningcited by 23
  5. 2022It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teachercited by 20
  6. 2022Genie: Show Me the Data for Quantizationcited by 17
  7. 2024CLAMP-ViT: Contrastive Data-Free Learning for Adaptive Post-Training Quantization of ViTscited by 17
  8. 2023Hard Sample Matters a Lot in Zero-Shot Quantizationcited by 15
  9. 2023Rethinking Data-Free Quantization as a Zero-Sum Gamecited by 13
  10. 2025Low-bit Model Quantization for Deep Neural Networks: A Surveycited by 11
  11. 2023Data-Free Quantization via Mixed-Precision Compensation without Fine-Tuningcited by 9
  12. 2024MimiQ: Low-Bit Data-Free Quantization of Vision Transformers with Encouraging Inter-Head Attention Similaritycited by 9
28 more
  1. 2023A Model for Every User and Budget: Label-Free and Personalized Mixed-Precision Quantizationcited by 8
  2. 2023GenQ: Quantization in Low Data Regimes with Generative Synthetic Datacited by 8
  3. 2025Zero-shot Quantization: A Comprehensive Surveycited by 8
  4. 2022Dynamic Data-Free Knowledge Distillation by Easy-to-Hard Learning Strategycited by 7
  5. 2024Dataset Size Recovery from LoRA Weightscited by 7
  6. 2023MBQuant: A Novel Multi-Branch Topology Method for Arbitrary Bit-width Network Quantizationcited by 6
  7. 2025Radio: Rate-Distortion Optimization for Large Language Model Compressioncited by 5
  8. 2022Long-Range Zero-Shot Generative Deep Network Quantizationcited by 4
  9. 2023Robustness-Guided Image Synthesis for Data-Free Quantizationcited by 4
  10. 2025Knowledge Distillation Detection for Open-weights Modelscited by 4
  11. 2022Towards Feature Distribution Alignment and Diversity Enhancement for Data-Free Quantizationcited by 3
  12. 2022Efficient Adaptive Activation Rounding for Post-Training Quantizationcited by 3
  13. 2022Redistribution of Weights and Activations for AdderNet Quantizationcited by 3
  14. 2024Semantic Alignment and Reinforcement for Data-Free Quantization of Vision Transformerscited by 3
  15. 2023ACQ: Improving Generative Data-free Quantization Via Attention Correctioncited by 2
  16. 2024Foundations of Large Language Model Compression -- Part 1: Weight Quantizationcited by 2
  17. 2024Data Generation for Hardware-Friendly Post-Training Quantizationcited by 2
  18. 2024Infrared Domain Adaptation with Zero-Shot Quantizationcited by 1
  19. 2024Advanced Knowledge Transfer: Refined Feature Distillation for Zero-Shot Quantization in Edge Computingcited by 1
  20. 2025Enhancing Generalization in Data-free Quantization via Mixup-class Promptingcited by 1
  21. 2026SynQ: Accurate Zero-shot Quantization by Synthesis-aware Fine-tuningcited by 1
  22. 2026LOCALUT: Harnessing Capacity-Computation Tradeoffs for LUT-Based Inference in DRAM-PIMcited by 1
  23. 2026Selective Coupling of Decoupled Informative Regions: Masked Attention Alignment for Data-Free Quantization of Vision Transformerscited by 1
  24. 2022Data-Free Quantization with Accurate Activation Clipping and Adaptive Batch Normalizationcited by 0
  25. 2022The Vanishing Decision Boundary Complexity and the Strong First Componentcited by 0
  26. 2025GranQ: Efficient Channel-wise Quantization via Vectorized Pre-Scaling for Zero-Shot QATcited by 0
  27. 2025Federated Domain Generalization with Latent Space Inversioncited by 0
  28. 2026Toward Multi-Domain and Long-Tailed Quantization via Feature Alignment and Scalingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2021Diverse Sample Generation: Pushing the Limit of Generative Data-free Quantization
  2. 2024Quantization without Tears
  3. 2021Quantization of Generative Adversarial Networks for Efficient Inference: a Methodological Study
  4. 2023Post-training Model Quantization Using GANs for Synthetic Data Generation
  5. 2024QGen: On the Ability to Generalize in Quantization Aware Training
  6. 2024A Data-Free Analytical Quantization Scheme for Deep Learning Models
  7. 2023MRQ:Support Multiple Quantization Schemes through Model Re-Quantization
  8. 2021MQBench: Towards Reproducible and Deployable Model Quantization Benchmark