Research map / Detection and segmentation / Open-set and few-shot detection
Information-Theoretic Segmentation by Inpainting Error Maximization
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 |
|---|---|---|---|
| Refine IEM masks using SegNet trained on high-error pseudo-labels | IEM · Flowers | IoU: 76.8 -> 78.9 | Helped |
| Refine IEM masks using SegNet trained on high-error pseudo-labels | IEM · CUB | IoU: 52.2 -> 55.1 | Helped |
| Refine IEM masks using SegNet trained on high-error pseudo-labels | IEM · LSUN Car | IoU: 65.1 -> 68.5 | Helped |
| Add foreground/background deviation regularization (ablation) | IEM · CUB | IoU: 50.6 -> 52.2 | Helped |
| Add foreground/background deviation regularization (ablation) | IEM · Flowers | IoU: 68.0 -> 76.8 | Helped |
| Smooth masks after updates (ablation) | IEM · CUB | IoU: 47.0 -> 52.2 | Helped |
| Restrict updates to mask-boundary pixels (ablation) | IEM · CUB | IoU: 42.8 -> 52.2 | Helped |
| Use the proposed L1 conditional-error objective rather than L2 (ablation) | IEM · CUB | IoU: 51.7 -> 52.2 | Helped |
| Use mask size rather than masked-image L1 norm for normalization (ablation) | IEM · CUB | IoU: 51.8 -> 52.2 | Helped |
| Use a simple Gaussian-filter inpainter instead of GatedConv (ablation) | IEM · CUB | IoU: 40.3 -> 52.2 | Helped |
| Use a simple Gaussian-filter inpainter instead of GatedConv with coarse outputs (ablation) | IEM · CUB | IoU: 41.6 -> 52.2 | Helped |
| Use a simple Gaussian-filter inpainter instead of GatedConv trained with L1 loss only (ablation) | IEM · CUB | IoU: 43.7 -> 52.2 | Helped |
Closest papers by meaning
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · similarity 0.86)
- Segment Anything without Supervision (NeurIPS 2024 · similarity 0.85)
- Learning Hierarchical Image Segmentation For Recognition and By Recognition (ICLR 2024 · similarity 0.85)
- Factorized Diffusion Architectures for Unsupervised Image Generation and Segmentation (NeurIPS 2024 · similarity 0.85)
- A Generative Approach at the Instance-Level for Image Segmentation Under Limited Training Data Conditions (Student Abstract) (AAAI 2025 · similarity 0.85)
- Latent Noise Segmentation: How Neural Noise Leads to the Emergence of Segmentation and Grouping (ICML 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.