STRIDE-4B
STRIDE (Structured Temporal Refinement with Iterative DEnoising) is a lightweight proactive activation model for streaming video understanding. It decides when a downstream Video-LLM should respond during a live video stream — without waiting for explicit user queries.
Paper: STRIDE: When to Speak Meets Sequence Denoising for Streaming Video Understanding
Junho Kim*, Hosu Lee*, James M. Rehg, Minsu Kim, Yong Man Ro
UIUC, KAIST, Google DeepMind
What is STRIDE?
Existing streaming Video-LLMs are reactive — they only respond when a user explicitly asks a question. STRIDE makes them proactive by adding a lightweight front-end that continuously monitors incoming frames and predicts coherent activation spans indicating when to trigger a response.
The key insight is that activation in streaming video is not a point-wise binary decision ("should I respond now?"), but a span-structured sequence modeling problem — the model must capture consistent onset (0 → 1), persistence (1 → 1), and offset (1 → 0) transitions. STRIDE achieves this through masked diffusion over a temporal activation window, jointly predicting and iteratively refining activation signals across the window.
Two-Stage Architecture
Video Stream
│
▼
[STRIDE Activation Model] ← this model (4B)
│
│ trigger (only if active)
▼
[Downstream Video-LLM] ← frozen, any off-the-shelf
│
▼
Response
- Stage 1 — Activation (STRIDE): Monitors the stream at 1 FPS, maintains a sliding activation window, and iteratively denoises binary activation labels via masked diffusion.
- Stage 2 — Response (Downstream LLM): When triggered, the frozen downstream Video-LLM receives the accumulated frame cache and generates a response. STRIDE is fully plug-and-play — compatible with any off-the-shelf Video-LLM.
Results
OVO-Bench (Online Video Understanding)
| Method | Real-Time Perception | Backward Tracing | Forward Active Responding | Overall |
|---|---|---|---|---|
| Flash-VStream-7B | 28.37 | 27.38 | 45.09 | 33.61 |
| Dispider | 54.55 | 36.06 | 34.72 | 41.78 |
| TimeChat-Online-7B | 58.60 | 42.00 | 36.40 | 45.60 |
| QueryStream-7B | 61.40 | 42.10 | 39.03 | 47.51 |
| StreamAgent-7B | 61.30 | 41.70 | 45.40 | 49.40 |
| STRIDE-4B + Gemma3-4B | 60.58 | 34.60 | 57.57 | 50.92 |
| STRIDE-4B + InternVL3-8B | 66.63 | 47.77 | 57.33 | 57.24 |
| STRIDE-4B + Qwen3-VL-8B | 69.77 | 48.67 | 62.37 | 60.27 |
StreamingBench (Streaming Comprehension)
| Method | Real-Time Visual | Omni-Source | Contextual | Overall |
|---|---|---|---|---|
| Flash-VStream-7B | 23.23 | 26.00 | 24.12 | 24.04 |
| VideoLLM-Online-8B | 35.99 | 28.45 | 26.55 | 32.48 |
| Dispider | 67.63 | 35.66 | 33.61 | 53.12 |
| StreamAgent-7B | 74.31 | 36.26 | 34.62 | 57.02 |
| STRIDE-4B + Gemma3-4B | 59.93 | 36.40 | 41.00 | 50.49 |
| STRIDE-4B + InternVL3-8B | 73.82 | 40.90 | 40.90 | 59.19 |
| STRIDE-4B + Qwen3-VL-8B | 76.01 | 40.00 | 39.90 | 59.98 |
Usage
For the full streaming inference pipeline and evaluation scripts, please refer to the STRIDE GitHub repository.
Training
- Architecture:
Qwen3VLForSTRIDE(Qwen3-VL backbone with a temporal activation head) - Base model: Qwen/Qwen3-VL-4B-Instruct
- Training data: Temporal activation annotations curated from eight publicly available video understanding datasets (ActivityNet-Captions, LITA, YouCook2, ET-Instruct, Charades, CharadesEgo, DiDeMo, Grounded-VideoLLM). STRIDE-4B is trained under the same data and training configuration as STRIDE-2B.
Model Variants
| Model | Params | Description |
|---|---|---|
| STRIDE-2B | 2B | Default activation model |
| STRIDE-4B (this) | 4B | Scaled variant with improved accuracy |
Citation
@article{kim2026stride,
title={STRIDE: When to Speak Meets Sequence Denoising for Streaming Video Understanding},
author={Kim, Junho and Lee, Hosu and Rehg, James M. and Kim, Minsu and Ro, Yong Man},
journal={arXiv preprint arXiv:2603.27593},
year={2026}
}
License
This model is released under the Apache 2.0 License.