Research map / Vision-language and multimodal / Human and embodied modeling
Deep Visual-Semantic Quantization for Efficient Image Retrieval
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 |
|---|---|---|---|
| use DVSQ instead of DQN | image retrieval · NUS-WIDE | MAP: 0.752 -> 0.797 | Helped |
| use DVSQ instead of DSH | image retrieval · CIFAR-10 | MAP: 0.659 -> 0.733 | Helped |
| use DVSQ instead of DQN | image retrieval · ImageNet | MAP: 0.625 -> 0.684 | Helped |
| jointly optimize visual-semantic embeddings and quantization rather than train them separately (ablation) | DVSQ · NUS-WIDE | MAP: 0.782 -> 0.797 | Helped |
| jointly optimize visual-semantic embeddings and quantization rather than train them separately (ablation) | DVSQ · CIFAR-10 | MAP: 0.705 -> 0.733 | Helped |
| jointly optimize visual-semantic embeddings and quantization rather than train them separately (ablation) | DVSQ · ImageNet | MAP: 0.671 -> 0.684 | Helped |
| use adaptive margin loss rather than softmax classification, with CQ on deep features (ablation) | DVSQ · NUS-WIDE | MAP: 0.676 -> 0.797 | Helped |
| use adaptive margin loss rather than softmax classification, with CQ on deep features (ablation) | DVSQ · CIFAR-10 | MAP: 0.634 -> 0.733 | Helped |
| use adaptive margin loss rather than softmax classification, with CQ on deep features (ablation) | DVSQ · ImageNet | MAP: 0.521 -> 0.684 | Helped |
| use dataset-adaptive margins rather than one fixed margin (ablation) | DVSQ · NUS-WIDE, CIFAR-10, and ImageNet | MAP | Helped |
| include quantization loss with a tuned positive trade-off rather than setting its weight to zero (ablation) | DVSQ · NUS-WIDE, CIFAR-10, and ImageNet | MAP | Mixed |
Closest papers by meaning
- Similarity Preserving Deep Asymmetric Quantization for Image Retrieval (AAAI 2019 · similarity 0.91)
- Deep Quantization Network for Efficient Image Retrieval (AAAI 2016 · similarity 0.89)
- Generalized Product Quantization Network for Semi-Supervised Image Retrieval (CVPR 2020 · similarity 0.88)
- HiHPQ: Hierarchical Hyperbolic Product Quantization for Unsupervised Image Retrieval (AAAI 2024 · similarity 0.87)
- QuARI: Query Adaptive Retrieval Improvement (NeurIPS 2025 · similarity 0.86)
- QSVD: Efficient Low-rank Approximation for Unified Query-Key-Value Weight Compression in Low-Precision Vision-Language Models (NeurIPS 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.