LampQ: Towards Accurate Layer-wise Mixed Precision Quantization for Vision Transformers
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (7) · 5691-5699
DOI: 10.1609/aaai.v40i7.37489 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Minjun Kim und 5 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref).
Abstract (Originalsprache)
How can we accurately quantize a pre-trained Vision Transformer model? Quantization algorithms compress Vision Transformers (ViTs) into low-bit formats, reducing memory and computation demands with minimal accuracy degradation. However, existing methods rely on uniform precision, ignoring the diverse sensitivity of ViT components to quantization. Metric-based Mixed Precision Quantization (MPQ) is a promising alternative, but previous MPQ methods for ViTs suffer from three major limitations: 1) coarse granularity, 2) mismatch in metric scale across component types, and 3) quantization-unaware bit allocation. In this paper, we propose LampQ (Layer-wise Mixed Precision Quantization for Vision Transformers), an accurate metric-based MPQ method for ViTs to overcome these limitations. LampQ performs layer-wise quantization to achieve both fine-grained control and efficient acceleration, incorporating a type-aware Fisher-based metric to measure sensitivity. Then, LampQ assigns bit-widths optimally through integer linear programming and further updates them iteratively. Extensive experiments show that LampQ provides the state-of-the-art performance in quantizing ViTs pre-trained on various tasks such as image classification, object detection, and zero-shot quantization.
Zitieren
APA
Kim, M., Lee, J., Kim, J., Yun, J., Kwon, Y., & Kang, U. (2026). LampQ: Towards Accurate Layer-wise Mixed Precision Quantization for Vision Transformers. Proceedings of the AAAI Conference on Artificial Intelligence, 40(7), 5691-5699. https://doi.org/10.1609/aaai.v40i7.37489
Vancouver
Kim M, Lee J, Kim J, Yun J, Kwon Y, Kang U. LampQ: Towards Accurate Layer-wise Mixed Precision Quantization for Vision Transformers. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(7):5691-5699. doi:10.1609/aaai.v40i7.37489
BibTeX
@article{Kim2026,
author = {Kim, Minjun and Lee, Jaeri and Kim, Jongjin and Yun, Jeongin and Kwon, Yongmo and Kang, U},
title = {LampQ: Towards Accurate Layer-wise Mixed Precision Quantization for Vision Transformers},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {7},
pages = {5691--5699},
doi = {10.1609/aaai.v40i7.37489}
}
EUVIMED