# RAVU: Retrieval Augmented Video Understanding with Compositional Reasoning over Graph

arXiv: https://arxiv.org/abs/2505.03173 · HTML: https://mlsanity.com/paper/2505.03173/

## Key facts

- Cited by 7 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 33 of its 39 references that are in the map.
- First posted on arXiv 6 May 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)
- Retrieval-augmented generation / [Multimodal RAG](https://mlsanity.com/rag/multimodal-rag/)

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

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actions](https://mlsanity.com/paper/2105.08276/) (2021)
- [EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding](https://mlsanity.com/paper/2308.09126/) (2023)
- [ViperGPT: Visual Inference via Python Execution for Reasoning](https://mlsanity.com/paper/2303.08128/) (2023)
- [MovieChat: From Dense Token to Sparse Memory for Long Video Understanding](https://mlsanity.com/paper/2307.16449/) (2023)
- [Learning to Compress Prompts with Gist Tokens](https://mlsanity.com/paper/2304.08467/) (2023)
- [LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding](https://mlsanity.com/paper/2410.17434/) (2024)
- [VideoAgent: Long-form Video Understanding with Large Language Model as Agent](https://mlsanity.com/paper/2403.10517/) (2024)
- [MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding](https://mlsanity.com/paper/2404.05726/) (2024)
- [Self-Chained Image-Language Model for Video Localization and Question Answering](https://mlsanity.com/paper/2305.06988/) (2023)
- [A Simple LLM Framework for Long-Range Video Question-Answering](https://mlsanity.com/paper/2312.17235/) (2023)
- [Video-of-Thought: Step-by-Step Video Reasoning from Perception to Cognition](https://mlsanity.com/paper/2501.03230/) (2024)
- [AssistGPT: A General Multi-modal Assistant that can Plan, Execute, Inspect, and Learn](https://mlsanity.com/paper/2306.08640/) (2023)
- [An Image Grid Can Be Worth a Video: Zero-shot Video Question Answering Using a VLM](https://mlsanity.com/paper/2403.18406/) (2024)
- [Video ReCap: Recursive Captioning of Hour-Long Videos](https://mlsanity.com/paper/2402.13250/) (2024)
- [Verbs in Action: Improving verb understanding in video-language models](https://mlsanity.com/paper/2304.06708/) (2023)
- [Scene Graph Generation: A Comprehensive Survey](https://mlsanity.com/paper/2201.00443/) (2022)
- [MovieChat+: Question-aware Sparse Memory for Long Video Question Answering](https://mlsanity.com/paper/2404.17176/) (2024)
- [HiTeA: Hierarchical Temporal-Aware Video-Language Pre-training](https://mlsanity.com/paper/2212.14546/) (2022)
- [VideoRAG: Retrieval-Augmented Generation over Video Corpus](https://mlsanity.com/paper/2501.05874/) (2025)
- [Panoptic Video Scene Graph Generation](https://mlsanity.com/paper/2311.17058/) (2023)
- [Memory Consolidation Enables Long-Context Video Understanding](https://mlsanity.com/paper/2402.05861/) (2024)
- [Zero-Shot Video Question Answering with Procedural Programs](https://mlsanity.com/paper/2312.00937/) (2023)

## Built on by (most cited first, 7 of 7)

- [SceneRAG: Scene-level Retrieval-Augmented Generation for Video Understanding](https://mlsanity.com/paper/2506.07600/) (2025)
- [Dynamic Hub-and-Spoke Memory for Streaming Video Understanding](https://mlsanity.com/paper/2608.30294/) (2026)
- [Reasoning Text-to-Video Retrieval via Digital Twin Video Representations and Large Language Models](https://mlsanity.com/paper/2511.12371/) (2025)
- [Spatiotemporal Knowledge Graphs as Persistent Scene Memory for Embodied Question Answering](https://mlsanity.com/paper/2510.01483/) (2025)
- [Rethinking RAG in Long Videos: What to Retrieve and How to Use It?](https://mlsanity.com/paper/2606.13141/) (2026)
- [Prompting-MammAlps: Fine-Grained Text-to-Video Retrieval for Camera-Trap Data](https://mlsanity.com/paper/2607.09876/) (2026)
- [CAM: Question Answering on Entity-Centric Videos with Continuous Extraction and Adaptive Querying](https://mlsanity.com/paper/2609.06504/) (2026)

## Abstract (arXiv)

Comprehending long videos remains a significant challenge for Large Multi-modal Models (LMMs). Current LMMs struggle to process even minutes to hours videos due to their lack of explicit memory and retrieval mechanisms. To address this limitation, we propose RAVU (Retrieval Augmented Video Understanding), a novel framework for video understanding enhanced by retrieval with compositional reasoning over a spatio-temporal graph. We construct a graph representation of the video, capturing both spatial and temporal relationships between entities. This graph serves as a long-term memory, allowing us to track objects and their actions across time. To answer complex queries, we decompose the queries into a sequence of reasoning steps and execute these steps on the graph, retrieving relevant key information. Our approach enables more accurate understanding of long videos, particularly for queries that require multi-hop reasoning and tracking objects across frames. Our approach demonstrate superior performances with limited retrieved frames (5-10) compared with other SOTA methods and baselines on two major video QA datasets, NExT-QA and EgoSchema.