AI Paper Atlas

Research map / Architectures and efficiency / Quantization

ViM-VQ: Efficient Post-Training Vector Quantization for Visual Mamba

Juncan Deng, Shuaiting Li, Zeyu Wang, Kedong Xu, Hong Gu, Kejie Huang

ICCV 2025 · Architectures and efficiency › Quantization · approach: Ultra-low-bit LLM quantization · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add fast convex combination optimization to basic VQ (ablation)ViM-VQ on Vim-T 2-bit · ImageNet-1K (2-bit weights)top-1 accuracy: 33.18 -> 71.42Improved
add incremental vector quantization on top of combination optimization (ablation)ViM-VQ on Vim-T 2-bit · ImageNet-1K (2-bit weights)top-1 accuracy: 71.42 -> 72.17Improved
increase candidate codewords n from 4 to 16 (ablation)ViM-VQ on Vim-T 2-bit · ImageNet-1K (2-bit weights)top-1 accuracy: 72.17 -> 72.31Improved
increase candidate codewords n from 16 to 64 (ablation)ViM-VQ on Vim-T 2-bit · ImageNet-1K (2-bit weights)top-1 accuracy: 72.31 -> 72.34No clear effect
use incremental quantization so inference matches calibration accuracyViM-VQ vs VQ4DiT on Vim-T 2-bit · ImageNet-1K (2-bit weights)calibration-to-inference gap: VQ4DiT 71.33 -> 71.04; ViM-VQ 72.17 -> 72.17Improved
use limited-candidate local convex combination instead of global soft assignment (DKM)ViM-VQ vs DKM · Vim-T 2-bit calibrationGPU memory: 25 GB -> 11 GBImproved
use customized CUDA kernel for matrix-vector multiplicationViM-VQ on Vim-B 2-bit · ImageNet-1K valinference time: 357s -> 328sImproved
apply ViM-VQ instead of DKM at 1-bitViM-VQ on VMamba-T · ADE20K (1-bit, single-scale)mIoU: 37.2 -> 41.1Improved
apply ViM-VQ instead of DKM at 1-bitViM-VQ on VMamba-T · MSCOCO (1-bit)box AP: 37.5 -> 40.9Improved
apply ViM-VQ instead of DKM at 3-bitViM-VQ on Vim-T · ImageNet-1K (3-bit weights)top-1 accuracy: 73.64 -> 74.79Improved
use vector quantization instead of uniform quantization at 2-bitViM-VQ family on Vim-T · ImageNet-1K (2-bit weights)top-1 accuracy: 0.16 -> 33.18Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy