arXiv Machine Learning July 9, 2026 By Seunghyeon Kim, Taesun Yeom, Jinho Kim, Wonpyo Park, Kyuyeun Kim, Jaeho Lee

Activation Quantization of Vision Encoders Needs Prefixing Registers

Read the original on arXiv Machine Learning →

arXiv:2510. 04547v5 Announce Type: replace Abstract: Large pretrained vision encoders are central to multimodal intelligence, powering applications from on-device vision processing to vision-language models.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv AI

Jun 9

I-Segmenter: Integer-Only Vision Transformer for Efficient Semantic Segmentation

arXiv:2509. 10334v2 Announce Type: replace-cross Abstract: Vision Transformers (ViTs) have recently achieved strong results in semantic segmentation, yet their deployment on resource-constrained devices remains limited due to their high memory footprint and computational cost.

By Jordan Sassoon, Michal Szczepanski, Martyna Poreba

llms computer-vision efficiency

arXiv Computation and Language

Sep 22

SalQ-VLM: Fine-Grained Saliency-Guided Quantization for Vision-Language Models

arXiv:2508.03351v3 Announce Type: replace-cross Abstract: Large language models (LLMs) have demonstrated remarkable capabilities across diverse language tasks, motivating their extension to vision-la...

By Yufei Xue, Yushi Huang, Lunjie Zhu, Jiawei Shao, Jun Zhang

llms efficiency multimodal benchmarks

arXiv Computer Vision

Sep 22

GPLQ: A General, Practical, and Lightning QAT Method for Vision Transformers

arXiv:2506.11784v2 Announce Type: replace Abstract: Vision Transformers (ViTs) are essential in computer vision but are computationally intensive, too. Model quantization, particularly to low bit-wid...

By Guang Liang, Xinyao Liu, Jianxin Wu

llms computer-vision efficiency

arXiv Machine Learning

Sep 24

RAMP: Robust Adaptive Mixed-Precision Quantization for Edge CPU Vision Models

The paper introduces RAMP, a method for robust adaptive mixed‑precision quantization of vision models on edge CPUs. It evaluates 13 sensitivity metrics across four neural networks, finding that Jensen‑Shannon Divergence consistently identifies layers that can be safely quantized. Using K‑Means clustering on these metrics, RAMP achieves near‑lossless accuracy with an average 1.81× speed‑up, while cautioning against excluding low‑speed‑up layers that can fragment the computational graph.

By David Poblaci\'on-Criado, Dario Garcia-Gasulla, Eduardo Quinones

efficiency

arXiv AI

Aug 28

PACE: A Unified Condense-and-Extract Paradigm for Fast VLM Inference

PACE introduces a training‑free Condense‑and‑Extract framework that speeds up Vision‑Language Model inference by first adaptively downsampling visual inputs before encoding and then selectively retaining essential tokens during decoding. The Adaptive Pixel Compressor (APC) reduces encoder workload while preserving global context, and the Dynamic Dual‑Attention Extractor (DDAE) keeps task‑critical details by fusing visual and language signals. Applied to Qwen2.5‑VL‑7B, PACE maintains 93.8% of performance using only 10% of visual tokens, achieving a 3.1× speedup in time to first token.

By Junjie Liu, Shengyuan Ye, Xu Chen

llms efficiency multimodal

arXiv Computer Vision

Sep 17

Decoder-Agnostic Token Merging for Vision Transformers: A Systematic Study of G2TM

The paper studies Graph-Guided Token Merging (G2TM), a module that reduces token count in Vision Transformers. It evaluates G2TM across multiple segmentation frameworks and decoder types, finding that its performance gains are tied to the encoder rather than the decoder. The authors report consistent reductions in GFLOPs (22‑47%) and throughput improvements (up to 74%) on ADE20K, with optimal hyperparameters depending mainly on backbone pre‑training and target dataset.

By Victor Bercy, Martyna Poreba, Michal Szczepanski, Samia Bouchafa

llms computer-vision efficiency benchmarks