Research map / Detection and segmentation / Open-set and few-shot detection
Adaptive Multi-Source Predictor for Zero-Shot Video Object Segmentation
Xiaoqi Zhao, Shijie Chang, Youwei Pang, Jiaxing Yang, Lihe Zhang, Huchuan Lu
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 depth, static saliency and optical flow to RGB input (ablation) | FPN · DAVIS16 val | mean J: 77.7 -> 82.3 | Improved |
| add optical flow to RGB input (ablation) | FPN · DAVIS16 val | mean J: 77.7 -> 80.2 | Improved |
| add ISAM to multi-source fusion (ablation) | AMSP · DAVIS16 val | mean J: 82.3 -> 83.1 | Improved |
| add MEM after ISAM (ablation) | AMSP · DAVIS16 val | mean J: 83.1 -> 85.1 | Improved |
| add FPM after MEM (ablation) | AMSP · DAVIS16 val | mean J: 85.1 -> 86.3 | Improved |
| add APF adaptive predictor fusion (ablation) | AMSP · DAVIS16 val | mean J: 86.3 -> 87.1 | Improved |
| add ISAM-like interaction to static predictor (ablation) | static predictor · DAVIS16 val | mean F: 75.8 -> 76.8 | Improved |
| add adaptive fusion over MOS alone (ablation) | AMSP · Youtube-Objects | mean J: 65.0 -> 75.0 | Improved |
| adaptive fusion instead of hard selection (ablation) | AMSP · Youtube-Objects | mean J: 70.0 -> 75.0 | Improved |
| MOS only versus SOS only with low-quality flow (ablation) | AMSP · Youtube-Objects | mean J: 74.0 -> 65.0 | Hurt |
| use ISAM and MEM instead of transformer modules (ablation) | AMSP · DAVIS16 val | mean F: 85.8 -> 87.3 | Improved |
| add APF to existing RGB-D SOD method (ablation) | RD3D · NJUD | F-measure: .923 -> .935 | Improved |
Closest papers by meaning
- Isomer: Isomerous Transformer for Zero-shot Video Object Segmentation (ICCV 2023 · similarity 0.88)
- Learning to Segment Moving Objects (IJCV 2018 · similarity 0.87)
- The Emergence of Objectness: Learning Zero-shot Segmentation from Videos (NeurIPS 2021 · similarity 0.87)
- MOVE: Motion-Guided Few-Shot Video Object Segmentation (ICCV 2025 · similarity 0.86)
- Segmenting Moving Objects via an Object-Centric Layered Representation (NeurIPS 2022 · similarity 0.86)
- GMOS: Grounding Moving Object Segmentation in 3D Space and Time (NeurIPS 2026 · similarity 0.85)
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.
