Research map / Architectures and efficiency / CNN design
Episodic CAMN: Contextual Attention-Based Memory Networks With Iterative Feedback for Scene Labeling
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 |
|---|---|---|---|
| Add one-episode contextual attention to FCN (ablation) | FCN · PASCAL-Context | IU (%): 35.11 -> 39.64 | Helped |
| Add one-episode contextual attention to FCN (ablation) | FCN · SIFT Flow | IU (%): 41.45 -> 43.88 | Helped |
| Add one-episode contextual attention to FCN (ablation) | FCN · PASCAL VOC 2011 validation | IU (%): 62.70 -> 66.91 | Helped |
| Add episodic memory feedback and iterative context refinement (ablation) | FCN+CAMN · PASCAL-Context | IU (%): 39.64 -> 41.18 | Helped |
| Add episodic memory feedback and iterative context refinement (ablation) | FCN+CAMN · SIFT Flow | IU (%): 43.88 -> 45.22 | Helped |
| Add episodic memory feedback and iterative context refinement (ablation) | FCN+CAMN · PASCAL VOC 2011 validation | IU (%): 66.91 -> 68.18 | Helped |
| Use Softmax rather than ReLU for attention weights (ablation) | Contextual Attention-based Network · Scene labeling | labeling performance | No clear effect |
| Increase recurrent context-refinement iterations beyond 5 (ablation) | Episodic-CAMN · Scene labeling | IU | Hurt |
| Use feed-forward relevance scores instead of concatenation, cosine similarity, or absolute distance (ablation) | Contextual Attention-based Network · Scene labeling | labeling performance | Helped |
Closest papers by meaning
- Contextual Convolutional Networks (ICLR 2023 · similarity 0.87)
- Recurrent Attentional Reinforcement Learning for Multi-Label Image Recognition (AAAI 2018 · similarity 0.86)
- Saliency-based Sequential Image Attention with Multiset Prediction (NeurIPS 2017 · similarity 0.86)
- Context Encoding for Semantic Segmentation (CVPR 2018 · similarity 0.86)
- Context-aware Attentional Pooling (CAP) for Fine-grained Visual Classification (AAAI 2021 · similarity 0.86)
- Multi-Path Feedback Recurrent Neural Networks for Scene Parsing (AAAI 2017 · 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.