# Decouple and Cache: KV Cache Construction for Streaming Video Understanding

arXiv: https://arxiv.org/abs/2605.01858 · HTML: https://mlsanity.com/paper/2605.01858/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Visual token pruning in multimodal models, one of 12 directions in Efficient LLM inference and serving.
- Builds on 33 of its 37 references that are in the map.
- First posted on arXiv 3 May 2026.

## Placed in

- Efficient LLM inference and serving / [Visual token pruning in multimodal models](https://mlsanity.com/inference/visual-token-pruning-in-multimodal-models/)

## Builds on (most cited first, 30 of 33)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [Ego4D: Around the World in 3,000 Hours of Egocentric Video](https://mlsanity.com/paper/2110.07058/) (2021)
- [Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation](https://mlsanity.com/paper/2108.12409/) (2021)
- [SnapKV: LLM Knows What You are Looking for Before Generation](https://mlsanity.com/paper/2404.14469/) (2024)
- [H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models](https://mlsanity.com/paper/2306.14048/) (2023)
- [Extending Context Window of Large Language Models via Positional Interpolation](https://mlsanity.com/paper/2306.15595/) (2023)
- [LMDrive: Closed-Loop End-to-End Driving with Large Language Models](https://mlsanity.com/paper/2312.07488/) (2023)
- [VideoLLM-online: Online Video Large Language Model for Streaming Video](https://mlsanity.com/paper/2406.11816/) (2024)
- [MovieNet: A Holistic Dataset for Movie Understanding](https://mlsanity.com/paper/2007.10937/) (2020)
- [Flash-VStream: Memory-Based Real-Time Understanding for Long Video Streams](https://mlsanity.com/paper/2406.08085/) (2024)
- [StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding](https://mlsanity.com/paper/2411.03628/) (2024)
- [OVO-Bench: How Far is Your Video-LLMs from Real-World Online Video Understanding?](https://mlsanity.com/paper/2501.05510/) (2025)
- [Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction](https://mlsanity.com/paper/2501.03218/) (2025)
- [LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models](https://mlsanity.com/paper/2308.16137/) (2023)
- [Streaming Video Question-Answering with In-context Video KV-Cache Retrieval](https://mlsanity.com/paper/2503.00540/) (2025)
- [TimeChat-Online: 80% Visual Tokens are Naturally Redundant in Streaming Videos](https://mlsanity.com/paper/2504.17343/) (2025)
- [StreamingVLM: Real-Time Understanding for Infinite Video Streams](https://mlsanity.com/paper/2510.09608/) (2025)
- [StreamBridge: Turning Your Offline Video Large Language Model into a Proactive Streaming Assistant](https://mlsanity.com/paper/2505.05467/) (2025)
- [LION-FS: Fast & Slow Video-Language Thinker as Online Video Assistant](https://mlsanity.com/paper/2503.03663/) (2025)
- [InfiniPot-V: Memory-Constrained KV Cache Compression for Streaming Video Understanding](https://mlsanity.com/paper/2506.15745/) (2025)
- [StreamMem: Query-Agnostic KV Cache Memory for Streaming Video Understanding](https://mlsanity.com/paper/2508.15717/) (2025)
- [VideoLLM-MoD: Efficient Video-Language Streaming with Mixture-of-Depths Vision Computation](https://mlsanity.com/paper/2408.16730/) (2024)
- [LiveVLM: Efficient Online Video Understanding via Streaming-Oriented KV Cache and Retrieval](https://mlsanity.com/paper/2505.15269/) (2025)
- [StreamMind: Unlocking Full Frame Rate Streaming Video Dialogue through Event-Gated Cognition](https://mlsanity.com/paper/2503.06220/) (2025)
- [Identifying and Mitigating Position Bias of Multi-image Vision-Language Models](https://mlsanity.com/paper/2503.13792/) (2025)
- [Memory-efficient Streaming VideoLLMs for Real-time Procedural Video Understanding](https://mlsanity.com/paper/2504.13915/) (2025)
- [Learning Pixel-Level Distinctions for Video Highlight Detection](https://mlsanity.com/paper/2204.04615/) (2022)

## Built on by (most cited first, 2 of 2)

- [Beyond Retrieval: Progressive Latent Memory Evolution for Streaming Video Understanding](https://mlsanity.com/paper/2609.04131/) (2026)
- [ChronoStitch: Training-Free Composition of Visual KV Memories for Long-Horizon Temporal Reasoning](https://mlsanity.com/paper/2607.19547/) (2026)

## Abstract (arXiv)

Streaming video understanding requires processing unbounded video streams with limited memory and computation, posing two key challenges. First, continuously constructing new and evicting old key-value(KV) caches is required for unbounded streams. Secondly, due to the high cost of collecting and training on unbounded streams, models must learn from short sequences while generalizing to long streams. Existing streaming VideoVLLMs fail to scale to unbounded video streams or focus on cache reuse strategies, leaving the impact of cache construction underexplored. In this paper, we propose Decoupled Streaming Cache(DSCache), a training-free cache construction mechanism that adapts pretrained offline models to streaming settings. DSCache maintains a cumulative past KV cache while constructing a separate instant cache on-demand, decoupled from past caches to preserve the informativeness of recent inputs. To enable position extrapolation beyond the training length, DSCache further incorporates a position-agnostic encoding strategy, ensuring KV caches to support unseen positions and preventing position overflow. Experiments on Streaming Video QA benchmarks demonstrate DSCache's state-of-the-art performance, with an average 2.5% accuracy gains over prior methods.