mlsanityData to 7 Oct 2026

RPIQ: Residual-Projected Multi-Collaboration Closed-Loop and Single Instance Quantization for Visually Impaired Assistance

arXiv 2601.02888 ↗First posted 6 Jan 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Visually impaired users face significant challenges in daily information access and real-time environmental perception, and there is an urgent need for intelligent assistive systems with accurate recognition capabilities. Although large-scale models provide effective solutions for perception and reasoning, their practical deployment on assistive devices is severely constrained by excessive memory consumption and high inference costs. Moreover, existing quantization strategies often ignore inter-block error accumulation, leading to degraded model stability. To address these challenges, this study proposes a novel quantization framework -- Residual-Projected Multi-Collaboration Closed-Loop and Single Instance Quantization(RPIQ), whose quantization process adopts a multi-collaborative closed-loop compensation scheme based on Single Instance Calibration and Gauss-Seidel Iterative Quantization. Experiments on various types of large-scale models, including language models such as OPT, Qwen, and LLaMA, as well as vision-language models such as CogVLM2, demonstrate that RPIQ can compress models to 4-bit representation while significantly reducing peak memory consumption (approximately 60%-75% reduction compared to original full-precision models). The method maintains performance highly close to full-precision models across multiple language and visual tasks, and exhibits excellent recognition and reasoning capabilities in key applications such as text understanding and visual question answering in complex scenarios. While verifying the effectiveness of RPIQ for deployment in real assistive systems, this study also advances the computational efficiency and reliability of large models, enabling them to provide visually impaired users with the required information accurately and rapidly.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 35 of its 46 references that are in the map. First posted on arXiv 6 Jan 2026.

Builds on

35 of 46 references.

References · 35

  1. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2025Qwen3 Technical Reportcited by 8,195
  4. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,464
  5. 2016Pointer Sentinel Mixture Modelscited by 2,992
  6. 2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,838
  7. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  8. 2023AWQ: Activation-aware Weight Quantization for LLM Compression and Accelerationcited by 1,234
  9. 2021A Survey of Quantization Methods for Efficient Neural Network Inferencecited by 798
  10. 2020Up or Down? Adaptive Rounding for Post-Training Quantizationcited by 600
  11. 2024QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMscited by 598
  12. 2022ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformerscited by 517
23 more
  1. 2021Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpuscited by 481
  2. 2021BRECQ: Pushing the Limit of Post-Training Quantization by Block Reconstructioncited by 464
  3. 2021A White Paper on Neural Network Quantizationcited by 453
  4. 2024VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Modelscited by 443
  5. 2018HAQ: Hardware-Aware Automated Quantization with Mixed Precisioncited by 426
  6. 2023OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Modelscited by 403
  7. 20218-bit Optimizers via Block-wise Quantizationcited by 395
  8. 2023SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compressioncited by 359
  9. 2024QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebookscited by 354
  10. 2019Data-Free Quantization Through Weight Equalization and Bias Correctioncited by 328
  11. 2023Atom: Low-bit Quantization for Efficient and Accurate LLM Servingcited by 255
  12. 2024CogVLM2: Visual Language Models for Image and Video Understandingcited by 214
  13. 2019SemEval-2017 Task 4: Sentiment Analysis in Twittercited by 209
  14. 2020Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programmingcited by 89
  15. 2025GPTAQ: Efficient Finetuning-Free Quantization for Asymmetric Calibrationcited by 54
  16. 2024VoCoT: Unleashing Visually Grounded Multi-Step Reasoning in Large Multi-Modal Modelscited by 42
  17. 2024LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detectioncited by 29
  18. 2024VIALM: A Survey and Benchmark of Visually Impaired Assistance with Large Modelscited by 23
  19. 2022RAPQ: Rescuing Accuracy for Power-of-Two Low-bit Post-training Quantizationcited by 14
  20. 2025Beyond Visual Perception: Insights from Smartphone Interaction of Visually Impaired Users with Large Multimodal Modelscited by 13
  21. 2024COMQ: A Backpropagation-Free Algorithm for Post-Training Quantizationcited by 12
  22. 2024AI-based Wearable Vision Assistance System for the Visually Impaired: Integrating Real-Time Object Recognition and Contextual Understanding Using Large Vision-Language Modelscited by 2
  23. 2025Scene-Aware Vectorized Memory Multi-Agent Framework with Cross-Modal Differentiated Quantization VLMs for Visually Impaired Assistancecited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026VEQ: Modality-Adaptive Quantization for MoE Vision-Language Models
  2. 2024MBQ: Modality-Balanced Quantization for Large Vision-Language Models
  3. 2025Less Redundancy: Boosting Practicality of Vision Language Model in Walking Assistants
  4. 2024@Bench: Benchmarking Vision-Language Models for Human-centered Assistive Technology
  5. 2024P4Q: Learning to Prompt for Quantization in Visual-language Models
  6. 2026Towards Understanding Best Practices for Quantization of Vision-Language Models
  7. 2026P4Q: Co-designing Token Pruning and Quantization for Vision-Language Model Acceleration
  8. 2026Quant Experts: Token-aware Adaptive Error Reconstruction with Mixture of Experts for Large Vision-Language Models Quantization