Research map / Video / Video segmentation and tracking
SSTVOS: Sparse Spatiotemporal Transformers for Video Object Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 sinusoidal positional encodings for X, Y, and T (ablation) | SST · DAVIS 2017 validation | J & F (%): $73\pm 2$ -> $75.6\pm 0.6$ | Helped |
| Increase temporal history from τ=1 to τ=2 (ablation) | SST · DAVIS 2017 validation | J & F (%): $75.8\pm 0.3$ -> $76.2\pm 0.3$ | Helped |
| Increase temporal history from τ=2 to τ=3 (ablation) | SST · DAVIS 2017 validation | J & F (%): $76.2\pm 0.3$ -> $76.5\pm 0.3$ | Helped |
| Increase grid-attention depth from 1 to 2 layers (ablation) | SST · DAVIS 2017 validation | J & F (%): 65.3 -> 66.1 | Helped |
| Increase grid-attention depth from 2 to 3 layers (ablation) | SST · DAVIS 2017 validation | J & F (%): 66.1 -> 64.2 | Hurt |
| Use local attention instead of grid attention (ablation) | SST · DAVIS 2017 validation | J & F (%): 66.1 -> 76.2 | Helped |
| Use strided attention instead of grid attention (ablation) | SST · DAVIS 2017 validation | J & F (%): 66.1 -> 69.5 | Helped |
| Use LocalStrided attention instead of grid attention (ablation) | SST · DAVIS 2017 validation | J & F (%): 66.1 -> 72.3 | Helped |
| Use sinusoidal positional encoding instead of no positional encoding (ablation) | SST · DAVIS 2017 validation | J & F (%): $73\pm 2$ -> $75.6\pm 0.6$ | Helped |
Closest papers by meaning
- Spatiotemporal CNN for Video Object Segmentation (CVPR 2019 · similarity 0.90)
- Recurrent Dynamic Embedding for Video Object Segmentation (CVPR 2022 · similarity 0.88)
- LiVOS: Light Video Object Segmentation with Gated Linear Matching (CVPR 2025 · similarity 0.87)
- Point-VOS: Pointing Up Video Object Segmentation (CVPR 2024 · similarity 0.86)
- Sparse Video-Gen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity (ICML 2025 · similarity 0.86)
- SAM-I2V: Upgrading SAM to Support Promptable Video Segmentation with Less than 0.2% Training Cost (CVPR 2025 · similarity 0.86)
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.