Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Task-Oriented Channel Attention for Fine-Grained Few-Shot Classification
SuBeen Lee, WonJun Moon, Hyun Seok Seong, Jae‐Pil Heo
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 SAM (support attention module) (ablation) | ProtoNet Conv-4 · CUB cropped 1-shot | accuracy: 62.90 -> 68.53 | Improved |
| add QAM to SAM (ablation) | ProtoNet Conv-4 · CUB cropped 1-shot | accuracy: 68.53 -> 69.94 | Improved |
| add IAM to SAM+QAM (full TDM+IAM) (ablation) | ProtoNet Conv-4 · CUB cropped 1-shot | accuracy: 69.94 -> 72.96 | Improved |
| add IAM alone (ablation) | ProtoNet Conv-4 · Aircraft 5-shot | accuracy: 68.96 -> 68.57 | Hurt |
| add SAM to IAM (ablation) | ProtoNet Conv-4 · Aircraft 5-shot | accuracy: 68.57 -> 70.67 | Improved |
| add TDM + IAM to CTX | CTX ResNet-12 · CUB cropped 1-shot | accuracy: 80.67 -> 83.76 | Improved |
| add TDM + IAM to ProtoNet | ProtoNet Conv-4 · meta-iNat 1-shot | accuracy: 55.37 -> 65.10 | Improved |
| add TDM to FRN | FRN Conv-4 · tiered meta-iNat 5-shot | accuracy: 63.45 -> 62.91 | Hurt |
| add IAM to FRN | FRN ResNet-12 · mini-ImageNet 1-shot | accuracy: 66.69 -> 66.96 | Improved |
| add IAM to Meta DeepBDC | Meta DeepBDC ResNet-12 · mini-ImageNet to CUB 5-shot | accuracy: 74.11 -> 77.71 | Improved |
| add TDM + IAM with cosine distance (ablation) | ProtoNet Conv-4 · CUB cropped 1-shot | accuracy: 68.69 -> 71.17 | Improved |
| use TDM + IAM instead of SENet (ablation) | ProtoNet Conv-4 · CUB cropped 1-shot | accuracy: 69.62 -> 72.96 | Improved |
Closest papers by meaning
- Task Discrepancy Maximization for Fine-Grained Few-Shot Classification (CVPR 2022 · similarity 0.97)
- On the Importance of Distractors for Few-Shot Classification (ICCV 2021 · similarity 0.87)
- Dual Attention Networks for Few-Shot Fine-Grained Recognition (AAAI 2022 · similarity 0.87)
- CAD: Co-Adapting Discriminative Features for Improved Few-Shot Classification (CVPR 2022 · similarity 0.86)
- Looking for the Devil in the Details: Learning Trilinear Attention Sampling Network for Fine-Grained Image Recognition (CVPR 2019 · similarity 0.86)
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.85)
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.
