Research map / Video / Video segmentation and tracking
HOPE: Hierarchical Object Prototype Encoding for Efficient Object Instance Search in Videos
Tan Yu, Yuwei Wu, Junsong Yuan
Check 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 |
|---|---|---|---|
| Use hierarchical object prototype encoding with spatial and temporal self-similarity | VGG-16 proposal features · Groundhog Day | mAP: 0.55 -> 0.70 | Improved |
| Use hierarchical object prototype encoding with spatial and temporal self-similarity | VGG-16 proposal features · Groundhog Day | mAP: 0.62 -> 0.70 | Improved |
| Use HOPE rather than exhaustive proposal storage | Groundhog Day · Groundhog Day | memory: 3.22 GB; 28.64 MB | Improved |
| Use temporally-constrained sphere k-means instead of sphere k-means (ablation) | HOPE dataset-level encoding · Groundhog Day and NTU-VOI | mAP | Improved |
| Use spatially-constrained sphere k-means instead of K-means (ablation) | HOPE frame-level encoding · Groundhog Day and NTU-VOI | mAP | Improved |
| Use spatially-constrained sphere k-means instead of sphere k-means (ablation) | HOPE frame-level encoding · Groundhog Day and NTU-VOI | mAP | Improved |
| Use frame-level object prototypes instead of K-mediods representatives (ablation) | HOPE frame-level encoding · Groundhog Day and NTU-VOI | mAP | Improved |
| Use frame-level object prototypes instead of SMRS representatives (ablation) | HOPE frame-level encoding · Groundhog Day and NTU-VOI | mAP | Improved |
| Use dataset-level object prototype encoding instead of sparse coding (ablation) | HOPE · Groundhog Day and NTU-VOI | mAP: 0.55 vs. 0.70 | Improved |
| Use dataset-level object prototype encoding instead of PQ (ablation) | HOPE · Groundhog Day | mAP: 0.62 vs. 0.70 | Improved |
| Use dataset-level object prototype encoding instead of PQ, OPQ, or α-PQ (ablation) | HOPE · Groundhog Day and NTU-VOI | mAP | Improved |
| Increase the number of nonzero elements in dataset-level codes (ablation) | HOPE · Groundhog Day | mAP | Improved |
Closest papers by meaning
- Compressive Quantization for Fast Object Instance Search in Videos (ICCV 2017 · similarity 0.88)
- Look Before You Match: Instance Understanding Matters in Video Object Segmentation (CVPR 2023 · similarity 0.86)
- How Can Objects Help Action Recognition? (CVPR 2023 · similarity 0.85)
- Semantics Meets Temporal Correspondence: Self-supervised Object-centric Learning in Videos (ICCV 2023 · similarity 0.85)
- Dataset-Driven Unsupervised Object Discovery for Region-Based Instance Image Retrieval (TPAMI 2022 · similarity 0.85)
- Temporally Consistent Object-Centric Learning by Contrasting Slots (CVPR 2025 · similarity 0.83)
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.
