Research map / Label-efficient and robust learning / Semi-supervised learning
Generalized Product Quantization Network for Semi-Supervised 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 GPQ rather than SSGAH | semi-supervised image retrieval · CIFAR-10 Protocol 1, 12-bits | mAP: 0.819 -> 0.858 | Helped |
| use GPQ rather than SSGAH | semi-supervised image retrieval · NUS-WIDE Protocol 1, 12-bits | mAP: 0.838 -> 0.852 | Helped |
| use GPQ rather than SSGAH | unseen-category retrieval · CIFAR-10 Protocol 2, 12-bits | mAP: 0.309 -> 0.321 | Helped |
| use GPQ rather than SSGAH | unseen-category retrieval · NUS-WIDE Protocol 2, 48-bits | mAP: 0.579 -> 0.586 | Helped |
| replace modified VGG with CNN-F (ablation) | GPQ · CIFAR-10 and NUS-WIDE Protocol 1 | mAP | Hurt |
| add classifier C, classification loss, and entropy regularization rather than removing C (ablation) | GPQ-H · CIFAR-10 and NUS-WIDE Protocol 1 | mAP | Helped |
| use N-pair Product Quantization loss rather than triplet loss (ablation) | GPQ · CIFAR-10 Protocol 1 | representation separation | Helped |
| balance classification and entropy-loss influences similarly, with both parameters set to 0.1 (ablation) | GPQ · Protocol 1, 12- and 48-bit codes | mAP | Helped |
Closest papers by meaning
- Similarity Preserving Deep Asymmetric Quantization for Image Retrieval (AAAI 2019 · similarity 0.89)
- Unleashing the Full Potential of Product Quantization for Large-Scale Image Retrieval (NeurIPS 2023 · similarity 0.89)
- Deep Visual-Semantic Quantization for Efficient Image Retrieval (CVPR 2017 · similarity 0.88)
- HiHPQ: Hierarchical Hyperbolic Product Quantization for Unsupervised Image Retrieval (AAAI 2024 · similarity 0.86)
- Generalized Debiased Semi-Supervised Hashing for Large-Scale Image Retrieval (AAAI 2025 · similarity 0.85)
- Learning from Loss Landscape: Generalizable Mixed-Precision Quantization via Adaptive Sharpness-Aware Gradient Aligning (ICML 2025 · similarity 0.84)
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.