Research
Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects
arXiv:2604.05546v2 Announce Type: replace Abstract: Large Vision-Language Models (LVLMs) enable sophisticated reasoning over images and videos, yet their inference is hindered by a systemic efficiency
arXiv:2604.05546v2 Announce Type: replace Abstract: Large Vision-Language Models (LVLMs) enable sophisticated reasoning over images and videos, yet their inference is hindered by a systemic efficiency barrier known as visual token dominance. This overhead is driven by a multi-regime interplay between high-resolution feature extraction, quadratic attention scaling, and memory bandwidth constraints. We present a systematic taxonomy of efficiency techniques structured around the inference lifecycle, consisting of encoding, prefilling, and decoding. Unlike prior reviews focused on isolated optimizations, we analyze the end-to-end pipeline to reveal how upstream decisions dictate downstream bottlenecks, covering compute-bound visual encoding, the intensive prefilling of massive contexts, and the ''visual memory wall'' in bandwidth-bound decoding. By decoupling the efficiency landscape into the axes of shaping information density, managing long-context attention, and overcoming memory limits, this work provides a structured analysis of how isolated optimizations compose to navigate the trade-off between visual fidelity and system efficiency. The survey concludes by outlining four future frontiers supported by pilot empirical insights, including hybrid compression based on functional unit sensitivity, modality-aware decoding with relaxed verification, progressive state management for streaming continuity, and stage-disaggregated serving through hardware-algorithm co-design. Our literature repository is at https://github.com/SuDIS-ZJU/Efficient-LVLMs-Inference.
Related
- Towards Efficient Large Vision-Language Models: A Comprehensive Survey on Inference Strategies
- See the Forest for the Trees: Loosely Speculative Decoding via Visual-Semantic Guidance for Efficient Inference of Video LLMs
- Growing a Multi-head Twig via Distillation and Reinforcement Learning to Accelerate Large Vision-Language Models
- YOCO++: Enhancing YOCO with KV Residual Connections for Efficient LLM Inference
- Sparse or Dense? A Mechanistic Estimation of Computation Density in Transformer-based LLMs
Source: arXiv cs.CL | 2026-04-15