REVIEW 25 cited by
StreamMem: Query-Agnostic KV Cache Memory for Streaming Video Understanding
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
StreamMem: Query-Agnostic KV Cache Memory for Streaming Video Understanding
read the original abstract
Multimodal large language models (MLLMs) have made significant progress in visual-language reasoning, but their ability to efficiently handle long videos remains limited. Despite recent advances in long-context MLLMs, storing and attending to the key-value (KV) cache for long visual contexts incurs substantial memory and computational overhead. Existing visual compression methods require either encoding the entire visual context before compression or having access to the questions in advance, which is impractical for long video understanding and multi-turn conversational settings. In this work, we propose StreamMem, a query-agnostic KV cache memory mechanism for streaming video understanding. Specifically, StreamMem encodes new video frames in a streaming manner, compressing the KV cache using attention scores between visual tokens and generic query tokens, while maintaining a fixed-size KV memory to enable efficient question answering (QA) in memory-constrained, long-video scenarios. Evaluation on three long video understanding and two streaming video question answering benchmarks shows that StreamMem achieves state-of-the-art performance in query-agnostic KV cache compression and is competitive with query-aware compression approaches.
Forward citations
Cited by 25 Pith papers
-
An Efficient Streaming Video Understanding Framework with Agentic Control
R3-Streaming uses cascaded control with age-aware memory forgetting and TB-GRPO reinforcement learning to reach SOTA scores of 57.92 on OVO-Bench and 76.36 on StreamingBench with 95-96% fewer visual tokens.
-
CoRDS: Coreset-based Representative and Diverse Selection for Streaming Video Understanding
CoRDS selects a compact KV-cache subset via joint-space coreset coverage and log-det diversity to outperform token-wise heuristics on long-video VLM benchmarks.
-
Semantic-Aware Adaptive Visual Memory for Streaming Video Understanding
SAVEMem improves streaming video understanding scores by adding semantic awareness to memory compression and query-adaptive retrieval without any model training.
-
Don't Pause! Every prediction matters in a streaming video
SPOT-Bench tests real-time streaming video perception with timeliness metrics, exposing limitations in current models and introducing AsynKV as an improved baseline.
-
Mosaic: Cross-Modal Clustering for Efficient Video Understanding
Mosaic uses cross-modal clusters as the unit for KVCache organization in VLMs to achieve up to 1.38x speedup in streaming long-video understanding.
-
STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction
STAC compresses KV caches in streaming 3D reconstruction transformers via temporal token preservation with decayed attention, spatial voxel compression, and chunked multi-frame optimization, delivering 10x memory redu...
-
ChronoStitch: Training-Free Composition of Visual KV Memories for Long-Horizon Temporal Reasoning
ChronoStitch re-bases stored video-chunk KV keys into a global multimodal RoPE frame and selectively recomputes a small slice of high-deviation tokens, recovering most of the joint-prefill temporal-reasoning gap at 3....
-
FOLIO: Focused Semantic Memory for Streaming Video Understanding
Entity-centered focus-guided streaming memory lifts Qwen3-VL-8B to 82.0/69.1 Perception/Backward on OVO-Bench and 74.5 on StreamingBench while cutting writer tokens by ~32%.
-
Kamera: Unified Position-Invariant Multimodal KV Cache for Training-Free Reuse
Kamera stores a low-rank patch with each position-free KV chunk to restore cross-chunk conditioning lost in naive reuse, enabling cheap reordering, sliding windows, and recall across attention mechanisms.
-
How Well Can Your Video Model Remember? Measuring Memory-Budget Trade-offs in Long Video Understanding
Fits a model where logit-accuracy scales linearly in log frame budget B with distance-dependent exponent α(D) that decays log-linearly with temporal distance D, based on 155k binary predictions across ten models.
-
FlowNar: Scalable Streaming Narration for Long-Form Videos
FlowNar achieves bounded memory and 3x higher throughput for streaming narration on Ego4D, EgoExo4D, and EpicKitchens100 by combining dynamic historical context removal with a Cross Linear Attentive Memory module.
-
An Efficient Streaming Video Understanding Framework with Agentic Control
R3-Streaming uses cascaded control, age-aware memory forgetting, and TB-GRPO reinforcement learning to reach SOTA scores on streaming video benchmarks while cutting visual token usage by 95-96%.
-
CodecSight: Leveraging Video Codec Signals for Efficient Streaming VLM Inference
CodecSight reuses video codec signals for online patch pruning before the vision transformer and selective KV-cache refresh in the LLM, delivering up to 3x higher throughput and 87% lower GPU compute than prior baseli...
-
Video Streaming Thinking: VideoLLMs Can Watch and Think Simultaneously
A 7B video model that generates intermediate text thoughts during playback, before the query arrives, improves streaming-video QA accuracy while keeping query-time latency near real-time.
-
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
HERMES organizes the KV cache into a hierarchical memory to enable real-time streaming video understanding in MLLMs, achieving 10x faster TTFT and up to 11.4% accuracy gains on streaming benchmarks with 68% fewer tokens.
-
LiveVLM: Efficient Online Video Understanding via Streaming-Oriented KV Cache and Retrieval
LiveVLM introduces VSB and PaR to compress and retrieve KV cache in streaming video LLMs, enabling LLaVA-OneVision to reach SOTA accuracy among training-free query-agnostic and training-based online models.
-
ViCoStream: Streaming VideoLLMs Can Run Beyond 100 FPS with Stage-Wise Coordinated Inference
ViCoStream is a new coordinated pipeline framework for streaming VideoLLMs that achieves 134 FPS video throughput and less than 50 ms TTFT on A100 while keeping accuracy near full-history baselines.
-
Linear Scaling Video VLMs for Long Video Understanding
StateKV is an inference-time technique that replaces quadratic self-attention prefill in video VLMs with a fixed-capacity importance-based recurrent state, keeping accuracy near full attention on long-video benchmarks...
-
OmniMem: Perturbation-aware Memory Compression for Streaming Audio-Visual LLMs
OmniMem achieves 2-4% higher accuracy than training-free baselines on long video benchmarks for audio-visual LLMs by using modality-aware KV cache allocation and perturbation-aware state selection, with further gains ...
-
MuKV: Multi-Grained KV Cache Compression for Long Streaming Video Question-Answering
MuKV adds multi-grained KV cache compression at patch-frame-segment levels plus semi-hierarchical retrieval to raise accuracy and cut memory in long video question-answering.
-
Decouple and Cache: KV Cache Construction for Streaming Video Understanding
DSCache decouples cumulative past and instant KV caches with position-agnostic encoding to adapt offline VideoVLLMs to streaming video, delivering 2.5% average accuracy gains on QA benchmarks.
-
cuRAMSES: Scalable AMR Optimizations for Large-Scale Cosmological Simulations
cuRAMSES replaces Hilbert-curve domain decomposition with recursive k-section partitioning and adds Morton-key hashing plus spatial binning to cut communication volume and accelerate feedback routines by up to 260x wh...
-
cuRAMSES: Scalable AMR Optimizations for Large-Scale Cosmological Simulations
Recursive k-section domain decomposition, Morton-key hashing, and GPU dispatch cut communication and memory bottlenecks in RAMSES while preserving conservation to ~0.5%.
-
Towards a Dynamic and Fixed-budget Memory Bank for Efficient Streaming Video Understanding
CausalMem constructs a dynamic fixed-budget memory bank for streaming video in MLLMs via online semantic basis updates, achieving 20x token compression and accuracy gains on benchmarks when applied to LLaVA-OneVision ...
-
Watch, Remember, Reason: Human-View Video Understanding with MLLMs
This is a survey that frames video MLLM research via a human-view formulation of perceptual representations, memory states, reasoning traces, and predictions, then reviews methods, datasets, benchmarks, and open problems.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.