mlsanityData to 7 Oct 2026

D4C: Data-Free Quantization for Contrastive Language-Image Pre-training Models

arXiv 2511.15411 ↗First posted 19 Nov 2025cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Data-Free Quantization (DFQ) offers a practical solution for model compression without requiring access to real data, making it particularly attractive in privacy-sensitive scenarios. While DFQ has shown promise for unimodal models, its extension to Vision-Language Models such as Contrastive Language-Image Pre-training (CLIP) models remains underexplored. In this work, we reveal that directly applying existing DFQ techniques to CLIP results in substantial performance degradation due to two key limitations: insufficient semantic content and low intra-image diversity in synthesized samples. To tackle these challenges, we propose D4C, the first DFQ framework tailored for CLIP. D4C synthesizes semantically rich and structurally diverse pseudo images through three key components: 1) Prompt-Guided Semantic Injection aligns generated images with real-world semantics using text prompts; 2) Structural Contrastive Generation reproduces compositional structures of natural images by leveraging foreground-background contrastive synthesis; and 3) Perturbation-Aware Enhancement applies controlled perturbations to improve sample diversity and robustness. These components jointly empower D4C to synthesize images that are both semantically informative and structurally diverse, effectively bridging the performance gap of DFQ on CLIP. Extensive experiments validate the effectiveness of D4C, showing significant performance improvements on various bit-widths and models.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 35 of its 47 references that are in the map. First posted on arXiv 19 Nov 2025.

Builds on

35 of 47 references.

References · 35

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  4. 2019Learned Step Size Quantizationcited by 604
  5. 2020Up or Down? Adaptive Rounding for Post-Training Quantizationcited by 600
  6. 2022MedCLIP: Contrastive Learning from Unpaired Medical Images and Textcited by 564
  7. 2021BRECQ: Pushing the Limit of Post-Training Quantization by Block Reconstructioncited by 464
  8. 2021A White Paper on Neural Network Quantizationcited by 453
  9. 2019Dreaming to Distill: Data-free Knowledge Transfer via DeepInversioncited by 419
  10. 2020ZeroQ: A Novel Zero Shot Quantization Frameworkcited by 265
  11. 2019Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networkscited by 234
  12. 2019Low-bit Quantization of Neural Networks for Efficient Inferencecited by 185
23 more
  1. 2022QDrop: Randomly Dropping Quantization for Extremely Low-bit Post-Training Quantizationcited by 175
  2. 2021PTQ4ViT: Post-training quantization for vision transformers with twin uniform quantizationcited by 173
  3. 2019Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networkscited by 143
  4. 2022RepQ-ViT: Scale Reparameterization for Post-Training Quantization of Vision Transformerscited by 134
  5. 2020Generative Low-bitwidth Data Free Quantizationcited by 87
  6. 2022PD-Quant: Post-Training Quantization based on Prediction Difference Metriccited by 73
  7. 2021Diversifying Sample Generation for Accurate Data-Free Quantizationcited by 63
  8. 2022Patch Similarity Aware Data-Free Quantization for Vision Transformerscited by 55
  9. 2023Oscillation-free Quantization for Low-bit Vision Transformerscited by 55
  10. 2021IntraQ: Learning Synthetic Images with Intra-Class Heterogeneity for Zero-Shot Network Quantizationcited by 43
  11. 2022PSAQ-ViT V2: Towards Accurate and General Data-Free Quantization for Vision Transformerscited by 42
  12. 2024MedCLIP-SAM: Bridging Text and Image Towards Universal Medical Image Segmentationcited by 37
  13. 2024LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detectioncited by 29
  14. 2024Towards Accurate Post-Training Quantization of Vision Transformers via Error Reductioncited by 23
  15. 2023Unified Data-Free Compression: Pruning and Quantization without Fine-Tuningcited by 22
  16. 2022Genie: Show Me the Data for Quantizationcited by 17
  17. 2024MimiQ: Low-Bit Data-Free Quantization of Vision Transformers with Encouraging Inter-Head Attention Similaritycited by 7
  18. 2024P4Q: Learning to Prompt for Quantization in Visual-language Modelscited by 7
  19. 2025Zero-shot Quantization: A Comprehensive Surveycited by 7
  20. 2024CLIP-RLDrive: Human-Aligned Autonomous Driving via CLIP-Based Reward Shaping in Reinforcement Learningcited by 5
  21. 2025Sharpness-Aware Data Generation for Zero-shot Quantizationcited by 5
  22. 2025Sparse Model Inversion: Efficient Inversion of Vision Transformers for Data-Free Applicationscited by 4
  23. 2025AHCQ-SAM: Toward Accurate and Hardware-Compatible Post-Training Segment Anything Model Quantizationcited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025HQ-CLIP: Leveraging Large Vision-Language Models to Create High-Quality Image-Text Datasets and CLIP Models
  2. 2025Post-pre-training for Modality Alignment in Vision-Language Foundation Models
  3. 2025TernaryCLIP: Efficiently Compressing Vision-Language Models with Ternary Weights and Distilled Knowledge
  4. 2024CLIP-CID: Efficient CLIP Distillation via Cluster-Instance Discrimination
  5. 2025CLIPTTA: Robust Contrastive Vision-Language Test-Time Adaptation
  6. 2026Sparse CLIP: Co-Optimizing Interpretability and Performance in Contrastive Learning
  7. 2024MLIP: Efficient Multi-Perspective Language-Image Pretraining with Exhaustive Data Utilization
  8. 2023A Closer Look at the Explainability of Contrastive Language-Image Pre-training