Research map / Video / Video-language understanding
SPARROW: Learning Spatial Precision and Temporal Referential Consistency in Pixel-Grounded Video MLLMs
Mohamad Alansari, Naufal Suryanto, Divya Velayudhan, Sajid Javed, Naoufel Werghi, Muzammal Naseer
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add SPARROW to VideoGLaMM | VideoGLaMM · MeViS val-u | J&F: 48.5 -> 57.4 (+8.9) | Improved |
| add SPARROW to UniPixel | UniPixel · MeViS val | J&F: 53.1 -> 54.4 (+1.3) | Improved |
| add SPARROW to GLUS | GLUS · MeViS val-u | J&F: 59.8 -> 61.9 (+0.3) | Improved |
| add SPARROW to VideoGLaMM | VideoGLaMM · Ref-DAVIS17 val | J&F: 69.5 -> 76.8 (+7.3) | Improved |
| add SPARROW to UniPixel | UniPixel · Ref-YTVOS val | J&F: 70.5 -> 70.7 (+0.2) | No clear effect |
| add SPARROW to VideoGLaMM | VideoGLaMM · VideoGCG | CLAIR: 28.2 -> 33.6 (+5.4) | Improved |
| add SPARROW to VideoGLaMM | VideoGLaMM · VideoGCG | mIoU: 62.34 -> 65.59 (+3.25) | Improved |
| add SPARROW to UniPixel | UniPixel · VidSTG interrogative | mIoU: 41.25 -> 46.74 (+5.49) | Improved |
| add SPARROW to GLUS | GLUS · VidSTG interrogative | mIoU: 29.92 -> 35.17 (+5.25) | Improved |
| train with TSF, no test-time TSF (ablation) | VideoGLaMM · Ref-DAVIS17 val | J&F: 69.5 -> 72.4 (+2.9) | Improved |
| add test-time TSF on top of train-only TSF (BOX off) (ablation) | SPARROW · Ref-DAVIS17 val | J&F: 72.4 -> 75.3 | Improved |
| enable BOX prompting without TSF (ablation) | VideoGLaMM · Ref-DAVIS17 val | J&F: 69.5 -> 72.5 (+3.0) | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- GLUS — GLUS: Global-Local Reasoning Unified into A Single Large Language Model for Video Segmentation (CVPR 2025) · better on 5 of 5 shared results
- VideoGLaMM — VideoGLaMM : A Large Multimodal Model for Pixel-Level Visual Grounding in Videos (CVPR 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models (CVPR 2025 · similarity 0.90)
- PixFoundation 2.0: Do Video Multi-Modal LLMs Use Motion in Visual Grounding? (NeurIPS 2026 · similarity 0.90)
- TimeSuite: Improving MLLMs for Long Video Understanding via Grounded Tuning (ICLR 2025 · similarity 0.89)
- Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis (CVPR 2025 · similarity 0.88)
- MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding (NeurIPS 2024 · similarity 0.88)
- Coarse Correspondences Boost Spatial-Temporal Reasoning in Multimodal Language Model (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
