# HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding

arXiv: https://arxiv.org/abs/2601.14724 · HTML: https://mlsanity.com/paper/2601.14724/

## Key facts

- Cited by 41 papers in the map (data through 14 Oct 2026).
- Placed in Video understanding with MLLMs, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 41 of its 55 references that are in the map.
- First posted on arXiv 21 Jan 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)
- 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 41)

- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Ego4D: Around the World in 3,000 Hours of Egocentric Video](https://mlsanity.com/paper/2110.07058/) (2021)
- [Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis](https://mlsanity.com/paper/2405.21075/) (2024)
- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [MVBench: A Comprehensive Multi-modal Video Understanding Benchmark](https://mlsanity.com/paper/2311.17005/) (2023)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs](https://mlsanity.com/paper/2406.07476/) (2024)
- [MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies](https://mlsanity.com/paper/2404.06395/) (2024)
- [VILA: On Pre-training for Visual Language Models](https://mlsanity.com/paper/2312.07533/) (2023)
- [NTU RGB+D: A Large Scale Dataset for 3D Human Activity Analysis](https://mlsanity.com/paper/1604.02808/) (2016)
- [EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding](https://mlsanity.com/paper/2308.09126/) (2023)
- [LLaVA-Video: Video Instruction Tuning With Synthetic Data](https://mlsanity.com/paper/2410.02713/) (2024)
- [Long Context Transfer from Language to Vision](https://mlsanity.com/paper/2406.16852/) (2024)
- [LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding](https://mlsanity.com/paper/2410.17434/) (2024)
- [Titans: Learning to Memorize at Test Time](https://mlsanity.com/paper/2501.00663/) (2024)
- [Perception Test: A Diagnostic Benchmark for Multimodal Video Models](https://mlsanity.com/paper/2305.13786/) (2023)
- [Memory in the Age of AI Agents](https://mlsanity.com/paper/2512.13564/) (2025)
- [VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos](https://mlsanity.com/paper/2405.19209/) (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)
- [Kangaroo: A Powerful Video-Language Model Supporting Long-context Video Input](https://mlsanity.com/paper/2408.15542/) (2024)
- [Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction](https://mlsanity.com/paper/2501.03218/) (2025)
- [Streaming Video Question-Answering with In-context Video KV-Cache Retrieval](https://mlsanity.com/paper/2503.00540/) (2025)
- [Video-CCAM: Enhancing Video-Language Understanding with Causal Cross-Attention Masks for Short and Long Videos](https://mlsanity.com/paper/2408.14023/) (2024)
- [TimeChat-Online: 80% Visual Tokens are Naturally Redundant in Streaming Videos](https://mlsanity.com/paper/2504.17343/) (2025)
- [DyCoke: Dynamic Compression of Tokens for Fast Video Large Language Models](https://mlsanity.com/paper/2411.15024/) (2024)
- [StreamForest: Efficient Online Video Understanding with Persistent Event Memory](https://mlsanity.com/paper/2509.24871/) (2025)

## Built on by (most cited first, 30 of 41)

- [A Simple Baseline for Streaming Video Understanding](https://mlsanity.com/paper/2604.02317/) (2026)
- [Thinking in Streaming Video](https://mlsanity.com/paper/2603.12938/) (2026)
- [JoyAI-VL-Interaction: Real-Time Vision-Language Interaction Intelligence](https://mlsanity.com/paper/2606.14777/) (2026)
- [EventMemAgent: Hierarchical Event-Centric Memory for Online Video Understanding with Adaptive Tool Use](https://mlsanity.com/paper/2602.15329/) (2026)
- [Semantic-Aware Adaptive Visual Memory for Streaming Video Understanding](https://mlsanity.com/paper/2605.07897/) (2026)
- [What Should a Streaming Video Model Remember?](https://mlsanity.com/paper/2606.16353/) (2026)
- [OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMs](https://mlsanity.com/paper/2606.03890/) (2026)
- [CurveStream: Boosting Streaming Video Understanding in MLLMs via Curvature-Aware Hierarchical Visual Memory Management](https://mlsanity.com/paper/2603.19571/) (2026)
- [CodecSight: Leveraging Video Codec Signals for Efficient Streaming VLM Inference](https://mlsanity.com/paper/2604.06036/) (2026)
- [EGOSTREAM: A Diagnostic Benchmark for Streaming Episodic Memory in Egocentric Vision](https://mlsanity.com/paper/2605.31557/) (2026)
- [ObjectStream: Latent Objects as Memory Anchors for Streaming Video Understanding](https://mlsanity.com/paper/2607.28312/) (2026)
- [StreamTTT: Reconciling Real-Time Perception and Long-Term Memory in Streaming VLMs](https://mlsanity.com/paper/2608.13416/) (2026)
- [OmniMem: Perturbation-aware Memory Compression for Streaming Audio-Visual LLMs](https://mlsanity.com/paper/2606.07577/) (2026)
- [Harnessing Streaming Video in the Wild](https://mlsanity.com/paper/2606.08615/) (2026)
- [FOLIO: Focused Semantic Memory for Streaming Video Understanding](https://mlsanity.com/paper/2607.13298/) (2026)
- [Kamera: Unified Position-Invariant Multimodal KV Cache for Training-Free Reuse](https://mlsanity.com/paper/2606.23581/) (2026)
- [StreamScout: Learning When to Look Deeper for Streaming Video Understanding](https://mlsanity.com/paper/2609.00291/) (2026)
- [StreamOPD: A Post-Training Recipe with Spatio-Temporal Cue Gating for Streaming Video Understanding](https://mlsanity.com/paper/2608.16320/) (2026)
- [ShallowStream: Index Shallow then Answer Deep for Streaming Video Understanding](https://mlsanity.com/paper/2609.02780/) (2026)
- [Linear Scaling Video VLMs for Long Video Understanding](https://mlsanity.com/paper/2605.31598/) (2026)
- [ViCoStream: Streaming VideoLLMs Can Run Beyond 100 FPS with Stage-Wise Coordinated Inference](https://mlsanity.com/paper/2606.19849/) (2026)
- [EgoPlay: Event-Triggered Video Editing for Egocentric Streams](https://mlsanity.com/paper/2607.24560/) (2026)
- [Aero Realtime: Fully Aligned Input-Output Streams for Low-Latency Streaming Multimodal Generation](https://mlsanity.com/paper/2608.08469/) (2026)
- [StreamFlow: Dynamic Memory Flows for Streaming Video Understanding](https://mlsanity.com/paper/2608.10949/) (2026)
- [Beyond Retrieval: Progressive Latent Memory Evolution for Streaming Video Understanding](https://mlsanity.com/paper/2609.04131/) (2026)
- [Omni-Streaming Thinking](https://mlsanity.com/paper/2609.15128/) (2026)
- [VLMaxxing through FrameMogging Training-Free Anti-Recomputation for Video Vision-Language Models](https://mlsanity.com/paper/2605.03351/) (2026)
- [HY-Himmel Technical Report: Hierarchical Interleaved Multi-stream Motion Encoding for Long Video Understanding](https://mlsanity.com/paper/2605.08158/) (2026)
- [Learning to See What You Need: Gaze Attention for Multimodal Large Language Models](https://mlsanity.com/paper/2605.13080/) (2026)
- [SLVMBench: Skill Learning from Video Memory](https://mlsanity.com/paper/2607.11312/) (2026)

## Abstract (arXiv)

Recent advancements in Multimodal Large Language Models (MLLMs) have demonstrated significant improvement in offline video understanding. However, extending these capabilities to streaming video inputs, remains challenging, as existing models struggle to simultaneously maintain stable understanding performance, real-time responses, and low GPU memory overhead. To address this challenge, we propose HERMES, a novel training-free architecture for real-time and accurate understanding of video streams. Based on a mechanistic attention investigation, we conceptualize KV cache as a hierarchical memory framework that encapsulates video information across multiple granularities. During inference, HERMES reuses a compact KV cache, enabling efficient streaming understanding under resource constraints. Notably, HERMES requires no auxiliary computations upon the arrival of user queries, thereby guaranteeing real-time responses for continuous video stream interactions, which achieves 10× faster TTFT compared to prior SOTA. Even when reducing video tokens by up to 68% compared with uniform sampling, HERMES achieves superior or comparable accuracy across all benchmarks, with up to 11.4% gains on streaming datasets.