Research map / Label-efficient and robust learning / Recognition and long-tail
Dual Directed Capsule Network for Very Low Resolution Image Recognition
Maneet Singh, Shruti Nagpal, Richa Singh, Mayank Vatsa
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 HR-anchor loss (ablation) | DirectCapsNet · SVHN VLR digits (8x8) | top-1 accuracy: 82.42 -> 84.51 | Improved |
| add HR-anchor loss (ablation) | DirectCapsNet · SVHN VLR digits (8x8) | top-5 accuracy: 90.15 -> 91.20 | Improved |
| add targeted reconstruction loss (ablation) | DirectCapsNet · SVHN VLR digits (8x8) | top-1 accuracy: 81.95 -> 84.51 | Improved |
| add targeted reconstruction loss (ablation) | DirectCapsNet · SVHN VLR digits (8x8) | top-5 accuracy: 90.35 -> 91.20 | Improved |
| add both directed losses to CapsNet trained on VLR images | CapsNet (VLR) · SVHN VLR digits (8x8) | top-1 accuracy: 79.19 -> 84.51 | Improved |
| replace CNN with capsule network on VLR images | CNN (VLR) · SVHN VLR digits (8x8) | top-1 accuracy: 45.29 -> 79.19 | Improved |
| train CapsNet on VLR images instead of HR images only | CapsNet · SVHN VLR digits (8x8) | top-1 accuracy: 77.82 -> 79.19 | Improved |
| use DirectCapsNet instead of SHSR | VLR face recognition · SVHN VLR digits (8x8) | rank-1 accuracy: 82.6 -> 94.5 | Improved |
| use DirectCapsNet instead of Centerloss for VLR | UCCS · SVHN VLR digits (8x8) | rank-1 accuracy: 93.40 -> 95.81 | Improved |
Closest papers by meaning
- Studying Very Low Resolution Recognition Using Deep Networks (CVPR 2016 · similarity 0.85)
- DeriveNet for (Very) Low Resolution Image Classification (TPAMI 2021 · similarity 0.84)
- OrthCaps: An Orthogonal CapsNet with Sparse Attention Routing and Pruning (CVPR 2024 · similarity 0.82)
- DeepCaps: Going Deeper With Capsule Networks (CVPR 2019 · similarity 0.82)
- STAR-Caps: Capsule Networks with Straight-Through Attentive Routing (NeurIPS 2019 · similarity 0.82)
- PT-CapsNet: A Novel Prediction-Tuning Capsule Network Suitable for Deeper Architectures (ICCV 2021 · similarity 0.81)
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.
